🔧 refactor(user_cache): remove unused JSON import

# Conflicts:
#	controller/channel.go
#	middleware/distributor.go
#	model/channel.go
#	model/user.go
#	model/user_cache.go
#	relay/common/relay_info.go

.env.example

@@ -7,6 +7,8 @@
 # 调试相关配置
 # 启用pprof
 # ENABLE_PPROF=true
+# 启用调试模式
+# DEBUG=true
 
 # 数据库相关配置
 # 数据库连接字符串
@@ -41,6 +43,14 @@
 # 更新任务启用
 # UPDATE_TASK=true
 
+# 对话超时设置
+# 所有请求超时时间，单位秒，默认为0，表示不限制
+# RELAY_TIMEOUT=0
+# 流模式无响应超时时间，单位秒，如果出现空补全可以尝试改为更大值
+# STREAMING_TIMEOUT=120
+
+# Gemini 识别图片 最大图片数量
+# GEMINI_VISION_MAX_IMAGE_NUM=16
 
 # 会话密钥
 # SESSION_SECRET=random_string
@@ -58,8 +68,6 @@
 # GET_MEDIA_TOKEN_NOT_STREAM=true
 # 设置 Dify 渠道是否输出工作流和节点信息到客户端
 # DIFY_DEBUG=true
-# 设置流式一次回复的超时时间
-# STREAMING_TIMEOUT=120
 
 
 # 节点类型

.github/PULL_REQUEST_TEMPLATE/pull_request_template.md

@@ -0,0 +1,19 @@
+### PR 类型
+
+- [ ] Bug 修复
+- [ ] 新功能
+- [ ] 文档更新
+- [ ] 其他
+
+### PR 是否包含破坏性更新？
+
+- [ ] 是
+- [ ] 否
+
+### PR 描述
+
+**请在下方详细描述您的 PR，包括目的、实现细节等。**
+
+### **重要提示**
+
+**所有 PR 都必须提交到 `alpha` 分支。请确保您的 PR 目标分支是 `alpha`。**

.github/workflows/macos-release.yml

@@ -26,6 +26,7 @@ jobs:
       - name: Build Frontend
         env:
           CI: ""
+          NODE_OPTIONS: "--max-old-space-size=4096"
         run: |
           cd web
           bun install

.github/workflows/pr-target-branch-check.yml

@@ -0,0 +1,21 @@
+name: Check PR Branching Strategy
+on:
+  pull_request:
+    types: [opened, synchronize, reopened, edited]
+
+jobs:
+  check-branching-strategy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Enforce branching strategy
+        run: |
+          if [[ "${{ github.base_ref }}" == "main" ]]; then
+            if [[ "${{ github.head_ref }}" != "alpha" ]]; then
+              echo "Error: Pull requests to 'main' are only allowed from the 'alpha' branch."
+              exit 1
+            fi
+          elif [[ "${{ github.base_ref }}" != "alpha" ]]; then
+            echo "Error: Pull requests must be targeted to the 'alpha' or 'main' branch."
+            exit 1
+          fi
+          echo "Branching strategy check passed."

README.md

@@ -27,9 +27,6 @@
   <a href="https://goreportcard.com/report/github.com/Calcium-Ion/new-api">
     <img src="https://goreportcard.com/badge/github.com/Calcium-Ion/new-api" alt="GoReportCard">
   </a>
-  <a href="https://coderabbit.ai">
-    <img src="https://img.shields.io/coderabbit/prs/github/QuantumNous/new-api?utm_source=oss&utm_medium=github&utm_campaign=QuantumNous%2Fnew-api&labelColor=171717&color=FF570A&link=https%3A%2F%2Fcoderabbit.ai&label=CodeRabbit+Reviews" alt="CodeRabbit Pull Request Reviews">
-  </a>
 </p>
 </div>
 

common/api_type.go

@@ -0,0 +1,71 @@
+package common
+
+import "one-api/constant"
+
+func ChannelType2APIType(channelType int) (int, bool) {
+	apiType := -1
+	switch channelType {
+	case constant.ChannelTypeOpenAI:
+		apiType = constant.APITypeOpenAI
+	case constant.ChannelTypeAnthropic:
+		apiType = constant.APITypeAnthropic
+	case constant.ChannelTypeBaidu:
+		apiType = constant.APITypeBaidu
+	case constant.ChannelTypePaLM:
+		apiType = constant.APITypePaLM
+	case constant.ChannelTypeZhipu:
+		apiType = constant.APITypeZhipu
+	case constant.ChannelTypeAli:
+		apiType = constant.APITypeAli
+	case constant.ChannelTypeXunfei:
+		apiType = constant.APITypeXunfei
+	case constant.ChannelTypeAIProxyLibrary:
+		apiType = constant.APITypeAIProxyLibrary
+	case constant.ChannelTypeTencent:
+		apiType = constant.APITypeTencent
+	case constant.ChannelTypeGemini:
+		apiType = constant.APITypeGemini
+	case constant.ChannelTypeZhipu_v4:
+		apiType = constant.APITypeZhipuV4
+	case constant.ChannelTypeOllama:
+		apiType = constant.APITypeOllama
+	case constant.ChannelTypePerplexity:
+		apiType = constant.APITypePerplexity
+	case constant.ChannelTypeAws:
+		apiType = constant.APITypeAws
+	case constant.ChannelTypeCohere:
+		apiType = constant.APITypeCohere
+	case constant.ChannelTypeDify:
+		apiType = constant.APITypeDify
+	case constant.ChannelTypeJina:
+		apiType = constant.APITypeJina
+	case constant.ChannelCloudflare:
+		apiType = constant.APITypeCloudflare
+	case constant.ChannelTypeSiliconFlow:
+		apiType = constant.APITypeSiliconFlow
+	case constant.ChannelTypeVertexAi:
+		apiType = constant.APITypeVertexAi
+	case constant.ChannelTypeMistral:
+		apiType = constant.APITypeMistral
+	case constant.ChannelTypeDeepSeek:
+		apiType = constant.APITypeDeepSeek
+	case constant.ChannelTypeMokaAI:
+		apiType = constant.APITypeMokaAI
+	case constant.ChannelTypeVolcEngine:
+		apiType = constant.APITypeVolcEngine
+	case constant.ChannelTypeBaiduV2:
+		apiType = constant.APITypeBaiduV2
+	case constant.ChannelTypeOpenRouter:
+		apiType = constant.APITypeOpenRouter
+	case constant.ChannelTypeXinference:
+		apiType = constant.APITypeXinference
+	case constant.ChannelTypeXai:
+		apiType = constant.APITypeXai
+	case constant.ChannelTypeCoze:
+		apiType = constant.APITypeCoze
+	}
+	if apiType == -1 {
+		return constant.APITypeOpenAI, false
+	}
+	return apiType, true
+}

common/constants.go

@@ -193,109 +193,3 @@ const (
 	ChannelStatusManuallyDisabled = 2 // also don't use 0
 	ChannelStatusAutoDisabled     = 3
 )
-
-const (
-	ChannelTypeUnknown        = 0
-	ChannelTypeOpenAI         = 1
-	ChannelTypeMidjourney     = 2
-	ChannelTypeAzure          = 3
-	ChannelTypeOllama         = 4
-	ChannelTypeMidjourneyPlus = 5
-	ChannelTypeOpenAIMax      = 6
-	ChannelTypeOhMyGPT        = 7
-	ChannelTypeCustom         = 8
-	ChannelTypeAILS           = 9
-	ChannelTypeAIProxy        = 10
-	ChannelTypePaLM           = 11
-	ChannelTypeAPI2GPT        = 12
-	ChannelTypeAIGC2D         = 13
-	ChannelTypeAnthropic      = 14
-	ChannelTypeBaidu          = 15
-	ChannelTypeZhipu          = 16
-	ChannelTypeAli            = 17
-	ChannelTypeXunfei         = 18
-	ChannelType360            = 19
-	ChannelTypeOpenRouter     = 20
-	ChannelTypeAIProxyLibrary = 21
-	ChannelTypeFastGPT        = 22
-	ChannelTypeTencent        = 23
-	ChannelTypeGemini         = 24
-	ChannelTypeMoonshot       = 25
-	ChannelTypeZhipu_v4       = 26
-	ChannelTypePerplexity     = 27
-	ChannelTypeLingYiWanWu    = 31
-	ChannelTypeAws            = 33
-	ChannelTypeCohere         = 34
-	ChannelTypeMiniMax        = 35
-	ChannelTypeSunoAPI        = 36
-	ChannelTypeDify           = 37
-	ChannelTypeJina           = 38
-	ChannelCloudflare         = 39
-	ChannelTypeSiliconFlow    = 40
-	ChannelTypeVertexAi       = 41
-	ChannelTypeMistral        = 42
-	ChannelTypeDeepSeek       = 43
-	ChannelTypeMokaAI         = 44
-	ChannelTypeVolcEngine     = 45
-	ChannelTypeBaiduV2        = 46
-	ChannelTypeXinference     = 47
-	ChannelTypeXai            = 48
-	ChannelTypeCoze           = 49
-	ChannelTypeKling          = 50
-	ChannelTypeDummy          // this one is only for count, do not add any channel after this
-
-)
-
-var ChannelBaseURLs = []string{
-	"",                                    // 0
-	"https://api.openai.com",              // 1
-	"https://oa.api2d.net",                // 2
-	"",                                    // 3
-	"http://localhost:11434",              // 4
-	"https://api.openai-sb.com",           // 5
-	"https://api.openaimax.com",           // 6
-	"https://api.ohmygpt.com",             // 7
-	"",                                    // 8
-	"https://api.caipacity.com",           // 9
-	"https://api.aiproxy.io",              // 10
-	"",                                    // 11
-	"https://api.api2gpt.com",             // 12
-	"https://api.aigc2d.com",              // 13
-	"https://api.anthropic.com",           // 14
-	"https://aip.baidubce.com",            // 15
-	"https://open.bigmodel.cn",            // 16
-	"https://dashscope.aliyuncs.com",      // 17
-	"",                                    // 18
-	"https://api.360.cn",                  // 19
-	"https://openrouter.ai/api",           // 20
-	"https://api.aiproxy.io",              // 21
-	"https://fastgpt.run/api/openapi",     // 22
-	"https://hunyuan.tencentcloudapi.com", //23
-	"https://generativelanguage.googleapis.com", //24
-	"https://api.moonshot.cn",                   //25
-	"https://open.bigmodel.cn",                  //26
-	"https://api.perplexity.ai",                 //27
-	"",                                          //28
-	"",                                          //29
-	"",                                          //30
-	"https://api.lingyiwanwu.com",               //31
-	"",                                          //32
-	"",                                          //33
-	"https://api.cohere.ai",                     //34
-	"https://api.minimax.chat",                  //35
-	"",                                          //36
-	"https://api.dify.ai",                       //37
-	"https://api.jina.ai",                       //38
-	"https://api.cloudflare.com",                //39
-	"https://api.siliconflow.cn",                //40
-	"",                                          //41
-	"https://api.mistral.ai",                    //42
-	"https://api.deepseek.com",                  //43
-	"https://api.moka.ai",                       //44
-	"https://ark.cn-beijing.volces.com",         //45
-	"https://qianfan.baidubce.com",              //46
-	"",                                          //47
-	"https://api.x.ai",                          //48
-	"https://api.coze.cn",                       //49
-	"https://api.klingai.com",                   //50
-}

common/endpoint_type.go

@@ -0,0 +1,41 @@
+package common
+
+import "one-api/constant"
+
+// GetEndpointTypesByChannelType 获取渠道最优先端点类型（所有的渠道都支持 OpenAI 端点）
+func GetEndpointTypesByChannelType(channelType int, modelName string) []constant.EndpointType {
+	var endpointTypes []constant.EndpointType
+	switch channelType {
+	case constant.ChannelTypeJina:
+		endpointTypes = []constant.EndpointType{constant.EndpointTypeJinaRerank}
+	//case constant.ChannelTypeMidjourney, constant.ChannelTypeMidjourneyPlus:
+	//	endpointTypes = []constant.EndpointType{constant.EndpointTypeMidjourney}
+	//case constant.ChannelTypeSunoAPI:
+	//	endpointTypes = []constant.EndpointType{constant.EndpointTypeSuno}
+	//case constant.ChannelTypeKling:
+	//	endpointTypes = []constant.EndpointType{constant.EndpointTypeKling}
+	//case constant.ChannelTypeJimeng:
+	//	endpointTypes = []constant.EndpointType{constant.EndpointTypeJimeng}
+	case constant.ChannelTypeAws:
+		fallthrough
+	case constant.ChannelTypeAnthropic:
+		endpointTypes = []constant.EndpointType{constant.EndpointTypeAnthropic, constant.EndpointTypeOpenAI}
+	case constant.ChannelTypeVertexAi:
+		fallthrough
+	case constant.ChannelTypeGemini:
+		endpointTypes = []constant.EndpointType{constant.EndpointTypeGemini, constant.EndpointTypeOpenAI}
+	case constant.ChannelTypeOpenRouter: // OpenRouter 只支持 OpenAI 端点
+		endpointTypes = []constant.EndpointType{constant.EndpointTypeOpenAI}
+	default:
+		if IsOpenAIResponseOnlyModel(modelName) {
+			endpointTypes = []constant.EndpointType{constant.EndpointTypeOpenAIResponse}
+		} else {
+			endpointTypes = []constant.EndpointType{constant.EndpointTypeOpenAI}
+		}
+	}
+	if IsImageGenerationModel(modelName) {
+		// add to first
+		endpointTypes = append([]constant.EndpointType{constant.EndpointTypeImageGeneration}, endpointTypes...)
+	}
+	return endpointTypes
+}

common/gin.go

@@ -2,10 +2,11 @@ package common
 
 import (
 	"bytes"
-	"encoding/json"
 	"github.com/gin-gonic/gin"
 	"io"
+	"one-api/constant"
 	"strings"
+	"time"
 )
 
 const KeyRequestBody = "key_request_body"
@@ -31,7 +32,7 @@ func UnmarshalBodyReusable(c *gin.Context, v any) error {
 	}
 	contentType := c.Request.Header.Get("Content-Type")
 	if strings.HasPrefix(contentType, "application/json") {
-		err = json.Unmarshal(requestBody, &v)
+		err = Unmarshal(requestBody, &v)
 	} else {
 		// skip for now
 		// TODO: someday non json request have variant model, we will need to implementation this
@@ -43,3 +44,45 @@ func UnmarshalBodyReusable(c *gin.Context, v any) error {
 	c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
 	return nil
 }
+
+func SetContextKey(c *gin.Context, key constant.ContextKey, value any) {
+	c.Set(string(key), value)
+}
+
+func GetContextKey(c *gin.Context, key constant.ContextKey) (any, bool) {
+	return c.Get(string(key))
+}
+
+func GetContextKeyString(c *gin.Context, key constant.ContextKey) string {
+	return c.GetString(string(key))
+}
+
+func GetContextKeyInt(c *gin.Context, key constant.ContextKey) int {
+	return c.GetInt(string(key))
+}
+
+func GetContextKeyBool(c *gin.Context, key constant.ContextKey) bool {
+	return c.GetBool(string(key))
+}
+
+func GetContextKeyStringSlice(c *gin.Context, key constant.ContextKey) []string {
+	return c.GetStringSlice(string(key))
+}
+
+func GetContextKeyStringMap(c *gin.Context, key constant.ContextKey) map[string]any {
+	return c.GetStringMap(string(key))
+}
+
+func GetContextKeyTime(c *gin.Context, key constant.ContextKey) time.Time {
+	return c.GetTime(string(key))
+}
+
+func GetContextKeyType[T any](c *gin.Context, key constant.ContextKey) (T, bool) {
+	if value, ok := c.Get(string(key)); ok {
+		if v, ok := value.(T); ok {
+			return v, true
+		}
+	}
+	var t T
+	return t, false
+}

common/http.go

@@ -0,0 +1,57 @@
+package common
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/gin-gonic/gin"
+)
+
+func CloseResponseBodyGracefully(httpResponse *http.Response) {
+	if httpResponse == nil || httpResponse.Body == nil {
+		return
+	}
+	err := httpResponse.Body.Close()
+	if err != nil {
+		SysError("failed to close response body: " + err.Error())
+	}
+}
+
+func IOCopyBytesGracefully(c *gin.Context, src *http.Response, data []byte) {
+	if c.Writer == nil {
+		return
+	}
+
+	body := io.NopCloser(bytes.NewBuffer(data))
+
+	// We shouldn't set the header before we parse the response body, because the parse part may fail.
+	// And then we will have to send an error response, but in this case, the header has already been set.
+	// So the httpClient will be confused by the response.
+	// For example, Postman will report error, and we cannot check the response at all.
+	if src != nil {
+		for k, v := range src.Header {
+			// avoid setting Content-Length
+			if k == "Content-Length" {
+				continue
+			}
+			c.Writer.Header().Set(k, v[0])
+		}
+	}
+
+	// set Content-Length header manually BEFORE calling WriteHeader
+	c.Writer.Header().Set("Content-Length", fmt.Sprintf("%d", len(data)))
+
+	// Write header with status code (this sends the headers)
+	if src != nil {
+		c.Writer.WriteHeader(src.StatusCode)
+	} else {
+		c.Writer.WriteHeader(http.StatusOK)
+	}
+
+	_, err := io.Copy(c.Writer, body)
+	if err != nil {
+		LogError(c, fmt.Sprintf("failed to copy response body: %s", err.Error()))
+	}
+}

common/init.go

@@ -4,6 +4,7 @@ import (
 	"flag"
 	"fmt"
 	"log"
+	"one-api/constant"
 	"os"
 	"path/filepath"
 	"strconv"
@@ -24,7 +25,7 @@ func printHelp() {
 	fmt.Println("Usage: one-api [--port <port>] [--log-dir <log directory>] [--version] [--help]")
 }
 
-func LoadEnv() {
+func InitEnv() {
 	flag.Parse()
 
 	if *PrintVersion {
@@ -95,4 +96,25 @@ func LoadEnv() {
 	GlobalWebRateLimitEnable = GetEnvOrDefaultBool("GLOBAL_WEB_RATE_LIMIT_ENABLE", true)
 	GlobalWebRateLimitNum = GetEnvOrDefault("GLOBAL_WEB_RATE_LIMIT", 60)
 	GlobalWebRateLimitDuration = int64(GetEnvOrDefault("GLOBAL_WEB_RATE_LIMIT_DURATION", 180))
+
+	initConstantEnv()
+}
+
+func initConstantEnv() {
+	constant.StreamingTimeout = GetEnvOrDefault("STREAMING_TIMEOUT", 120)
+	constant.DifyDebug = GetEnvOrDefaultBool("DIFY_DEBUG", true)
+	constant.MaxFileDownloadMB = GetEnvOrDefault("MAX_FILE_DOWNLOAD_MB", 20)
+	// ForceStreamOption 覆盖请求参数，强制返回usage信息
+	constant.ForceStreamOption = GetEnvOrDefaultBool("FORCE_STREAM_OPTION", true)
+	constant.GetMediaToken = GetEnvOrDefaultBool("GET_MEDIA_TOKEN", true)
+	constant.GetMediaTokenNotStream = GetEnvOrDefaultBool("GET_MEDIA_TOKEN_NOT_STREAM", true)
+	constant.UpdateTask = GetEnvOrDefaultBool("UPDATE_TASK", true)
+	constant.AzureDefaultAPIVersion = GetEnvOrDefaultString("AZURE_DEFAULT_API_VERSION", "2025-04-01-preview")
+	constant.GeminiVisionMaxImageNum = GetEnvOrDefault("GEMINI_VISION_MAX_IMAGE_NUM", 16)
+	constant.NotifyLimitCount = GetEnvOrDefault("NOTIFY_LIMIT_COUNT", 2)
+	constant.NotificationLimitDurationMinute = GetEnvOrDefault("NOTIFICATION_LIMIT_DURATION_MINUTE", 10)
+	// GenerateDefaultToken 是否生成初始令牌，默认关闭。
+	constant.GenerateDefaultToken = GetEnvOrDefaultBool("GENERATE_DEFAULT_TOKEN", false)
+	// 是否启用错误日志
+	constant.ErrorLogEnabled = GetEnvOrDefaultBool("ERROR_LOG_ENABLED", false)
 }

common/json.go

@@ -5,14 +5,18 @@ import (
 	"encoding/json"
 )
 
-func DecodeJson(data []byte, v any) error {
-	return json.NewDecoder(bytes.NewReader(data)).Decode(v)
+func Unmarshal(data []byte, v any) error {
+	return json.Unmarshal(data, v)
 }
 
-func DecodeJsonStr(data string, v any) error {
-	return DecodeJson(StringToByteSlice(data), v)
+func UnmarshalJsonStr(data string, v any) error {
+	return json.Unmarshal(StringToByteSlice(data), v)
 }
 
-func EncodeJson(v any) ([]byte, error) {
+func DecodeJson(reader *bytes.Reader, v any) error {
+	return json.NewDecoder(reader).Decode(v)
+}
+
+func Marshal(v any) ([]byte, error) {
 	return json.Marshal(v)
 }

common/model.go

@@ -0,0 +1,42 @@
+package common
+
+import "strings"
+
+var (
+	// OpenAIResponseOnlyModels is a list of models that are only available for OpenAI responses.
+	OpenAIResponseOnlyModels = []string{
+		"o3-pro",
+		"o3-deep-research",
+		"o4-mini-deep-research",
+	}
+	ImageGenerationModels = []string{
+		"dall-e-3",
+		"dall-e-2",
+		"gpt-image-1",
+		"prefix:imagen-",
+		"flux-",
+		"flux.1-",
+	}
+)
+
+func IsOpenAIResponseOnlyModel(modelName string) bool {
+	for _, m := range OpenAIResponseOnlyModels {
+		if strings.Contains(modelName, m) {
+			return true
+		}
+	}
+	return false
+}
+
+func IsImageGenerationModel(modelName string) bool {
+	modelName = strings.ToLower(modelName)
+	for _, m := range ImageGenerationModels {
+		if strings.Contains(modelName, m) {
+			return true
+		}
+		if strings.HasPrefix(m, "prefix:") && strings.HasPrefix(modelName, strings.TrimPrefix(m, "prefix:")) {
+			return true
+		}
+	}
+	return false
+}

common/page_info.go

@@ -0,0 +1,62 @@
+package common
+
+import (
+	"github.com/gin-gonic/gin"
+	"strconv"
+)
+
+type PageInfo struct {
+	Page           int   `json:"page"`            // page num 页码
+	PageSize       int   `json:"page_size"`       // page size 页大小
+	StartTimestamp int64 `json:"start_timestamp"` // 秒级
+	EndTimestamp   int64 `json:"end_timestamp"`   // 秒级
+
+	Total int `json:"total"` // 总条数，后设置
+	Items any `json:"items"` // 数据，后设置
+}
+
+func (p *PageInfo) GetStartIdx() int {
+	return (p.Page - 1) * p.PageSize
+}
+
+func (p *PageInfo) GetEndIdx() int {
+	return p.Page * p.PageSize
+}
+
+func (p *PageInfo) GetPageSize() int {
+	return p.PageSize
+}
+
+func (p *PageInfo) GetPage() int {
+	return p.Page
+}
+
+func (p *PageInfo) SetTotal(total int) {
+	p.Total = total
+}
+
+func (p *PageInfo) SetItems(items any) {
+	p.Items = items
+}
+
+func GetPageQuery(c *gin.Context) (*PageInfo, error) {
+	pageInfo := &PageInfo{}
+	err := c.BindQuery(pageInfo)
+	if err != nil {
+		return nil, err
+	}
+	if pageInfo.Page < 1 {
+		// 兼容
+		page, _ := strconv.Atoi(c.Query("p"))
+		if page != 0 {
+			pageInfo.Page = page
+		} else {
+			pageInfo.Page = 1
+		}
+	}
+
+	if pageInfo.PageSize == 0 {
+		pageInfo.PageSize = ItemsPerPage
+	}
+	return pageInfo, nil
+}

common/redis.go

@@ -16,6 +16,10 @@ import (
 var RDB *redis.Client
 var RedisEnabled = true
 
+func RedisKeyCacheSeconds() int {
+	return SyncFrequency
+}
+
 // InitRedisClient This function is called after init()
 func InitRedisClient() (err error) {
 	if os.Getenv("REDIS_CONN_STRING") == "" {

common/str.go

@@ -1,6 +1,7 @@
 package common
 
 import (
+	"encoding/base64"
 	"encoding/json"
 	"math/rand"
 	"strconv"
@@ -31,16 +32,30 @@ func MapToJsonStr(m map[string]interface{}) string {
 	return string(bytes)
 }
 
-func StrToMap(str string) map[string]interface{} {
+func StrToMap(str string) (map[string]interface{}, error) {
 	m := make(map[string]interface{})
-	err := json.Unmarshal([]byte(str), &m)
+	err := Unmarshal([]byte(str), &m)
 	if err != nil {
-		return nil
+		return nil, err
 	}
-	return m
+	return m, nil
 }
 
-func IsJsonStr(str string) bool {
+func StrToJsonArray(str string) ([]interface{}, error) {
+	var js []interface{}
+	err := json.Unmarshal([]byte(str), &js)
+	if err != nil {
+		return nil, err
+	}
+	return js, nil
+}
+
+func IsJsonArray(str string) bool {
+	var js []interface{}
+	return json.Unmarshal([]byte(str), &js) == nil
+}
+
+func IsJsonObject(str string) bool {
 	var js map[string]interface{}
 	return json.Unmarshal([]byte(str), &js) == nil
 }
@@ -68,3 +83,15 @@ func StringToByteSlice(s string) []byte {
 	tmp2 := [3]uintptr{tmp1[0], tmp1[1], tmp1[1]}
 	return *(*[]byte)(unsafe.Pointer(&tmp2))
 }
+
+func EncodeBase64(str string) string {
+	return base64.StdEncoding.EncodeToString([]byte(str))
+}
+
+func GetJsonString(data any) string {
+	if data == nil {
+		return ""
+	}
+	b, _ := json.Marshal(data)
+	return string(b)
+}

constant/README.md

@@ -0,0 +1,26 @@
+# constant 包 (`/constant`)
+
+该目录仅用于放置全局可复用的**常量定义**，不包含任何业务逻辑或依赖关系。
+
+## 当前文件
+
+| 文件                   | 说明                                                                  |
+|----------------------|---------------------------------------------------------------------|
+| `azure.go`           | 定义与 Azure 相关的全局常量，如 `AzureNoRemoveDotTime`（控制删除 `.` 的截止时间）。         |
+| `cache_key.go`       | 缓存键格式字符串及 Token 相关字段常量，统一缓存命名规则。                                    |
+| `channel_setting.go` | Channel 级别的设置键，如 `proxy`、`force_format` 等。                          |
+| `context_key.go`     | 定义 `ContextKey` 类型以及在整个项目中使用的上下文键常量（请求时间、Token/Channel/User 相关信息等）。 |
+| `env.go`             | 环境配置相关的全局变量，在启动阶段根据配置文件或环境变量注入。                                     |
+| `finish_reason.go`   | OpenAI/GPT 请求返回的 `finish_reason` 字符串常量集合。                           |
+| `midjourney.go`      | Midjourney 相关错误码及动作(Action)常量与模型到动作的映射表。                            |
+| `setup.go`           | 标识项目是否已完成初始化安装 (`Setup` 布尔值)。                                       |
+| `task.go`            | 各种任务(Task)平台、动作常量及模型与动作映射表，如 Suno、Midjourney 等。                     |
+| `user_setting.go`    | 用户设置相关键常量以及通知类型(Email/Webhook)等。                                    |
+
+## 使用约定
+
+1. `constant` 包**只能被其他包引用**（import），**禁止在此包中引用项目内的其他自定义包**。如确有需要，仅允许引用 **Go 标准库**。
+2. 不允许在此目录内编写任何与业务流程、数据库操作、第三方服务调用等相关的逻辑代码。
+3. 新增类型时，请保持命名语义清晰，并在本 README 的 **当前文件** 表格中补充说明，确保团队成员能够快速了解其用途。
+
+> ⚠️ 违反以上约定将导致包之间产生不必要的耦合，影响代码可维护性与可测试性。请在提交代码前自行检查。

constant/api_type.go

@@ -0,0 +1,34 @@
+package constant
+
+const (
+	APITypeOpenAI = iota
+	APITypeAnthropic
+	APITypePaLM
+	APITypeBaidu
+	APITypeZhipu
+	APITypeAli
+	APITypeXunfei
+	APITypeAIProxyLibrary
+	APITypeTencent
+	APITypeGemini
+	APITypeZhipuV4
+	APITypeOllama
+	APITypePerplexity
+	APITypeAws
+	APITypeCohere
+	APITypeDify
+	APITypeJina
+	APITypeCloudflare
+	APITypeSiliconFlow
+	APITypeVertexAi
+	APITypeMistral
+	APITypeDeepSeek
+	APITypeMokaAI
+	APITypeVolcEngine
+	APITypeBaiduV2
+	APITypeOpenRouter
+	APITypeXinference
+	APITypeXai
+	APITypeCoze
+	APITypeDummy // this one is only for count, do not add any channel after this
+)

constant/cache_key.go

@@ -1,12 +1,5 @@
 package constant
 
-import "one-api/common"
-
-// 使用函数来避免初始化顺序带来的赋值问题
-func RedisKeyCacheSeconds() int {
-	return common.SyncFrequency
-}
-
 // Cache keys
 const (
 	UserGroupKeyFmt    = "user_group:%d"

constant/channel.go

@@ -0,0 +1,109 @@
+package constant
+
+const (
+	ChannelTypeUnknown        = 0
+	ChannelTypeOpenAI         = 1
+	ChannelTypeMidjourney     = 2
+	ChannelTypeAzure          = 3
+	ChannelTypeOllama         = 4
+	ChannelTypeMidjourneyPlus = 5
+	ChannelTypeOpenAIMax      = 6
+	ChannelTypeOhMyGPT        = 7
+	ChannelTypeCustom         = 8
+	ChannelTypeAILS           = 9
+	ChannelTypeAIProxy        = 10
+	ChannelTypePaLM           = 11
+	ChannelTypeAPI2GPT        = 12
+	ChannelTypeAIGC2D         = 13
+	ChannelTypeAnthropic      = 14
+	ChannelTypeBaidu          = 15
+	ChannelTypeZhipu          = 16
+	ChannelTypeAli            = 17
+	ChannelTypeXunfei         = 18
+	ChannelType360            = 19
+	ChannelTypeOpenRouter     = 20
+	ChannelTypeAIProxyLibrary = 21
+	ChannelTypeFastGPT        = 22
+	ChannelTypeTencent        = 23
+	ChannelTypeGemini         = 24
+	ChannelTypeMoonshot       = 25
+	ChannelTypeZhipu_v4       = 26
+	ChannelTypePerplexity     = 27
+	ChannelTypeLingYiWanWu    = 31
+	ChannelTypeAws            = 33
+	ChannelTypeCohere         = 34
+	ChannelTypeMiniMax        = 35
+	ChannelTypeSunoAPI        = 36
+	ChannelTypeDify           = 37
+	ChannelTypeJina           = 38
+	ChannelCloudflare         = 39
+	ChannelTypeSiliconFlow    = 40
+	ChannelTypeVertexAi       = 41
+	ChannelTypeMistral        = 42
+	ChannelTypeDeepSeek       = 43
+	ChannelTypeMokaAI         = 44
+	ChannelTypeVolcEngine     = 45
+	ChannelTypeBaiduV2        = 46
+	ChannelTypeXinference     = 47
+	ChannelTypeXai            = 48
+	ChannelTypeCoze           = 49
+	ChannelTypeKling          = 50
+	ChannelTypeJimeng         = 51
+	ChannelTypeDummy          // this one is only for count, do not add any channel after this
+
+)
+
+var ChannelBaseURLs = []string{
+	"",                                    // 0
+	"https://api.openai.com",              // 1
+	"https://oa.api2d.net",                // 2
+	"",                                    // 3
+	"http://localhost:11434",              // 4
+	"https://api.openai-sb.com",           // 5
+	"https://api.openaimax.com",           // 6
+	"https://api.ohmygpt.com",             // 7
+	"",                                    // 8
+	"https://api.caipacity.com",           // 9
+	"https://api.aiproxy.io",              // 10
+	"",                                    // 11
+	"https://api.api2gpt.com",             // 12
+	"https://api.aigc2d.com",              // 13
+	"https://api.anthropic.com",           // 14
+	"https://aip.baidubce.com",            // 15
+	"https://open.bigmodel.cn",            // 16
+	"https://dashscope.aliyuncs.com",      // 17
+	"",                                    // 18
+	"https://api.360.cn",                  // 19
+	"https://openrouter.ai/api",           // 20
+	"https://api.aiproxy.io",              // 21
+	"https://fastgpt.run/api/openapi",     // 22
+	"https://hunyuan.tencentcloudapi.com", //23
+	"https://generativelanguage.googleapis.com", //24
+	"https://api.moonshot.cn",                   //25
+	"https://open.bigmodel.cn",                  //26
+	"https://api.perplexity.ai",                 //27
+	"",                                          //28
+	"",                                          //29
+	"",                                          //30
+	"https://api.lingyiwanwu.com",               //31
+	"",                                          //32
+	"",                                          //33
+	"https://api.cohere.ai",                     //34
+	"https://api.minimax.chat",                  //35
+	"",                                          //36
+	"https://api.dify.ai",                       //37
+	"https://api.jina.ai",                       //38
+	"https://api.cloudflare.com",                //39
+	"https://api.siliconflow.cn",                //40
+	"",                                          //41
+	"https://api.mistral.ai",                    //42
+	"https://api.deepseek.com",                  //43
+	"https://api.moka.ai",                       //44
+	"https://ark.cn-beijing.volces.com",         //45
+	"https://qianfan.baidubce.com",              //46
+	"",                                          //47
+	"https://api.x.ai",                          //48
+	"https://api.coze.cn",                       //49
+	"https://api.klingai.com",                   //50
+	"https://visual.volcengineapi.com",          //51
+}

constant/channel_setting.go

@@ -1,7 +0,0 @@
-package constant
-
-var (
-	ForceFormat                     = "force_format"        // ForceFormat 强制格式化为OpenAI格式
-	ChanelSettingProxy              = "proxy"               // Proxy 代理
-	ChannelSettingThinkingToContent = "thinking_to_content" // ThinkingToContent
-)

constant/context_key.go

@@ -1,11 +1,42 @@
 package constant
 
+type ContextKey string
+
 const (
-	ContextKeyRequestStartTime = "request_start_time"
-	ContextKeyUserSetting      = "user_setting"
-	ContextKeyUserQuota        = "user_quota"
-	ContextKeyUserStatus       = "user_status"
-	ContextKeyUserEmail        = "user_email"
-	ContextKeyUserGroup        = "user_group"
-	ContextKeyUsingGroup       = "group"
+	ContextKeyOriginalModel    ContextKey = "original_model"
+	ContextKeyRequestStartTime ContextKey = "request_start_time"
+
+	/* token related keys */
+	ContextKeyTokenUnlimited         ContextKey = "token_unlimited_quota"
+	ContextKeyTokenKey               ContextKey = "token_key"
+	ContextKeyTokenId                ContextKey = "token_id"
+	ContextKeyTokenGroup             ContextKey = "token_group"
+	ContextKeyTokenAllowIps          ContextKey = "allow_ips"
+	ContextKeyTokenSpecificChannelId ContextKey = "specific_channel_id"
+	ContextKeyTokenModelLimitEnabled ContextKey = "token_model_limit_enabled"
+	ContextKeyTokenModelLimit        ContextKey = "token_model_limit"
+
+	/* channel related keys */
+	ContextKeyChannelId                ContextKey = "channel_id"
+	ContextKeyChannelName              ContextKey = "channel_name"
+	ContextKeyChannelCreateTime        ContextKey = "channel_create_name"
+	ContextKeyChannelBaseUrl           ContextKey = "base_url"
+	ContextKeyChannelType              ContextKey = "channel_type"
+	ContextKeyChannelSetting           ContextKey = "channel_setting"
+	ContextKeyChannelParamOverride     ContextKey = "param_override"
+	ContextKeyChannelOrganization      ContextKey = "channel_organization"
+	ContextKeyChannelAutoBan           ContextKey = "auto_ban"
+	ContextKeyChannelModelMapping      ContextKey = "model_mapping"
+	ContextKeyChannelStatusCodeMapping ContextKey = "status_code_mapping"
+	ContextKeyChannelIsMultiKey        ContextKey = "channel_is_multi_key"
+
+	/* user related keys */
+	ContextKeyUserId      ContextKey = "id"
+	ContextKeyUserSetting ContextKey = "user_setting"
+	ContextKeyUserQuota   ContextKey = "user_quota"
+	ContextKeyUserStatus  ContextKey = "user_status"
+	ContextKeyUserEmail   ContextKey = "user_email"
+	ContextKeyUserGroup   ContextKey = "user_group"
+	ContextKeyUsingGroup  ContextKey = "group"
+	ContextKeyUserName    ContextKey = "username"
 )

constant/endpoint_type.go

@@ -0,0 +1,16 @@
+package constant
+
+type EndpointType string
+
+const (
+	EndpointTypeOpenAI          EndpointType = "openai"
+	EndpointTypeOpenAIResponse  EndpointType = "openai-response"
+	EndpointTypeAnthropic       EndpointType = "anthropic"
+	EndpointTypeGemini          EndpointType = "gemini"
+	EndpointTypeJinaRerank      EndpointType = "jina-rerank"
+	EndpointTypeImageGeneration EndpointType = "image-generation"
+	//EndpointTypeMidjourney     EndpointType = "midjourney-proxy"
+	//EndpointTypeSuno           EndpointType = "suno-proxy"
+	//EndpointTypeKling          EndpointType = "kling"
+	//EndpointTypeJimeng         EndpointType = "jimeng"
+)

constant/env.go

@@ -1,9 +1,5 @@
 package constant
 
-import (
-	"one-api/common"
-)
-
 var StreamingTimeout int
 var DifyDebug bool
 var MaxFileDownloadMB int
@@ -17,39 +13,3 @@ var NotifyLimitCount int
 var NotificationLimitDurationMinute int
 var GenerateDefaultToken bool
 var ErrorLogEnabled bool
-
-//var GeminiModelMap = map[string]string{
-//	"gemini-1.0-pro": "v1",
-//}
-
-func InitEnv() {
-	StreamingTimeout = common.GetEnvOrDefault("STREAMING_TIMEOUT", 120)
-	DifyDebug = common.GetEnvOrDefaultBool("DIFY_DEBUG", true)
-	MaxFileDownloadMB = common.GetEnvOrDefault("MAX_FILE_DOWNLOAD_MB", 20)
-	// ForceStreamOption 覆盖请求参数，强制返回usage信息
-	ForceStreamOption = common.GetEnvOrDefaultBool("FORCE_STREAM_OPTION", true)
-	GetMediaToken = common.GetEnvOrDefaultBool("GET_MEDIA_TOKEN", true)
-	GetMediaTokenNotStream = common.GetEnvOrDefaultBool("GET_MEDIA_TOKEN_NOT_STREAM", true)
-	UpdateTask = common.GetEnvOrDefaultBool("UPDATE_TASK", true)
-	AzureDefaultAPIVersion = common.GetEnvOrDefaultString("AZURE_DEFAULT_API_VERSION", "2025-04-01-preview")
-	GeminiVisionMaxImageNum = common.GetEnvOrDefault("GEMINI_VISION_MAX_IMAGE_NUM", 16)
-	NotifyLimitCount = common.GetEnvOrDefault("NOTIFY_LIMIT_COUNT", 2)
-	NotificationLimitDurationMinute = common.GetEnvOrDefault("NOTIFICATION_LIMIT_DURATION_MINUTE", 10)
-	// GenerateDefaultToken 是否生成初始令牌，默认关闭。
-	GenerateDefaultToken = common.GetEnvOrDefaultBool("GENERATE_DEFAULT_TOKEN", false)
-	// 是否启用错误日志
-	ErrorLogEnabled = common.GetEnvOrDefaultBool("ERROR_LOG_ENABLED", false)
-
-	//modelVersionMapStr := strings.TrimSpace(os.Getenv("GEMINI_MODEL_MAP"))
-	//if modelVersionMapStr == "" {
-	//	return
-	//}
-	//for _, pair := range strings.Split(modelVersionMapStr, ",") {
-	//	parts := strings.Split(pair, ":")
-	//	if len(parts) == 2 {
-	//		GeminiModelMap[parts[0]] = parts[1]
-	//	} else {
-	//		common.SysError(fmt.Sprintf("invalid model version map: %s", pair))
-	//	}
-	//}
-}

constant/midjourney.go

@@ -22,6 +22,8 @@ const (
 	MjActionPan           = "PAN"
 	MjActionSwapFace      = "SWAP_FACE"
 	MjActionUpload        = "UPLOAD"
+	MjActionVideo         = "VIDEO"
+	MjActionEdits         = "EDITS"
 )
 
 var MidjourneyModel2Action = map[string]string{
@@ -41,4 +43,6 @@ var MidjourneyModel2Action = map[string]string{
 	"mj_pan":            MjActionPan,
 	"swap_face":         MjActionSwapFace,
 	"mj_upload":         MjActionUpload,
+	"mj_video":          MjActionVideo,
+	"mj_edits":          MjActionEdits,
 }

constant/multi_key_mode.go

@@ -0,0 +1,8 @@
+package constant
+
+type MultiKeyMode string
+
+const (
+	MultiKeyModeRandom  MultiKeyMode = "random"  // 随机
+	MultiKeyModePolling MultiKeyMode = "polling" // 轮询
+)

constant/task.go

@@ -6,11 +6,15 @@ const (
 	TaskPlatformSuno       TaskPlatform = "suno"
 	TaskPlatformMidjourney              = "mj"
 	TaskPlatformKling      TaskPlatform = "kling"
+	TaskPlatformJimeng     TaskPlatform = "jimeng"
 )
 
 const (
 	SunoActionMusic  = "MUSIC"
 	SunoActionLyrics = "LYRICS"
+
+	TaskActionGenerate     = "generate"
+	TaskActionTextGenerate = "textGenerate"
 )
 
 var SunoModel2Action = map[string]string{

constant/user_setting.go

@@ -1,16 +0,0 @@
-package constant
-
-var (
-	UserSettingNotifyType            = "notify_type"                    // QuotaWarningType 额度预警类型
-	UserSettingQuotaWarningThreshold = "quota_warning_threshold"        // QuotaWarningThreshold 额度预警阈值
-	UserSettingWebhookUrl            = "webhook_url"                    // WebhookUrl webhook地址
-	UserSettingWebhookSecret         = "webhook_secret"                 // WebhookSecret webhook密钥
-	UserSettingNotificationEmail     = "notification_email"             // NotificationEmail 通知邮箱地址
-	UserAcceptUnsetRatioModel        = "accept_unset_model_ratio_model" // AcceptUnsetRatioModel 是否接受未设置价格的模型
-	UserSettingRecordIpLog          = "record_ip_log"                   // 是否记录请求和错误日志IP
-)
-
-var (
-	NotifyTypeEmail   = "email"   // Email 邮件
-	NotifyTypeWebhook = "webhook" // Webhook
-)

controller/channel-billing.go

@@ -8,6 +8,7 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/model"
 	"one-api/service"
 	"one-api/setting"
@@ -341,34 +342,34 @@ func updateChannelMoonshotBalance(channel *model.Channel) (float64, error) {
 }
 
 func updateChannelBalance(channel *model.Channel) (float64, error) {
-	baseURL := common.ChannelBaseURLs[channel.Type]
+	baseURL := constant.ChannelBaseURLs[channel.Type]
 	if channel.GetBaseURL() == "" {
 		channel.BaseURL = &baseURL
 	}
 	switch channel.Type {
-	case common.ChannelTypeOpenAI:
+	case constant.ChannelTypeOpenAI:
 		if channel.GetBaseURL() != "" {
 			baseURL = channel.GetBaseURL()
 		}
-	case common.ChannelTypeAzure:
+	case constant.ChannelTypeAzure:
 		return 0, errors.New("尚未实现")
-	case common.ChannelTypeCustom:
+	case constant.ChannelTypeCustom:
 		baseURL = channel.GetBaseURL()
 	//case common.ChannelTypeOpenAISB:
 	//	return updateChannelOpenAISBBalance(channel)
-	case common.ChannelTypeAIProxy:
+	case constant.ChannelTypeAIProxy:
 		return updateChannelAIProxyBalance(channel)
-	case common.ChannelTypeAPI2GPT:
+	case constant.ChannelTypeAPI2GPT:
 		return updateChannelAPI2GPTBalance(channel)
-	case common.ChannelTypeAIGC2D:
+	case constant.ChannelTypeAIGC2D:
 		return updateChannelAIGC2DBalance(channel)
-	case common.ChannelTypeSiliconFlow:
+	case constant.ChannelTypeSiliconFlow:
 		return updateChannelSiliconFlowBalance(channel)
-	case common.ChannelTypeDeepSeek:
+	case constant.ChannelTypeDeepSeek:
 		return updateChannelDeepSeekBalance(channel)
-	case common.ChannelTypeOpenRouter:
+	case constant.ChannelTypeOpenRouter:
 		return updateChannelOpenRouterBalance(channel)
-	case common.ChannelTypeMoonshot:
+	case constant.ChannelTypeMoonshot:
 		return updateChannelMoonshotBalance(channel)
 	default:
 		return 0, errors.New("尚未实现")

controller/channel-test.go

@@ -11,14 +11,15 @@ import (
 	"net/http/httptest"
 	"net/url"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/middleware"
 	"one-api/model"
 	"one-api/relay"
 	relaycommon "one-api/relay/common"
-	"one-api/relay/constant"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strconv"
 	"strings"
 	"sync"
@@ -29,20 +30,23 @@ import (
 	"github.com/gin-gonic/gin"
 )
 
-func testChannel(channel *model.Channel, testModel string) (err error, openAIErrorWithStatusCode *dto.OpenAIErrorWithStatusCode) {
+func testChannel(channel *model.Channel, testModel string) (err error, newAPIError *types.NewAPIError) {
 	tik := time.Now()
-	if channel.Type == common.ChannelTypeMidjourney {
+	if channel.Type == constant.ChannelTypeMidjourney {
 		return errors.New("midjourney channel test is not supported"), nil
 	}
-	if channel.Type == common.ChannelTypeMidjourneyPlus {
-		return errors.New("midjourney plus channel test is not supported!!!"), nil
+	if channel.Type == constant.ChannelTypeMidjourneyPlus {
+		return errors.New("midjourney plus channel test is not supported"), nil
 	}
-	if channel.Type == common.ChannelTypeSunoAPI {
+	if channel.Type == constant.ChannelTypeSunoAPI {
 		return errors.New("suno channel test is not supported"), nil
 	}
-	if channel.Type == common.ChannelTypeKling {
+	if channel.Type == constant.ChannelTypeKling {
 		return errors.New("kling channel test is not supported"), nil
 	}
+	if channel.Type == constant.ChannelTypeJimeng {
+		return errors.New("jimeng channel test is not supported"), nil
+	}
 	w := httptest.NewRecorder()
 	c, _ := gin.CreateTestContext(w)
 
@@ -53,7 +57,7 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 		strings.HasPrefix(testModel, "m3e") || // m3e 系列模型
 		strings.Contains(testModel, "bge-") || // bge 系列模型
 		strings.Contains(testModel, "embed") ||
-		channel.Type == common.ChannelTypeMokaAI { // 其他 embedding 模型
+		channel.Type == constant.ChannelTypeMokaAI { // 其他 embedding 模型
 		requestPath = "/v1/embeddings" // 修改请求路径
 	}
 
@@ -95,14 +99,14 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 
 	err = helper.ModelMappedHelper(c, info, nil)
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeChannelModelMappedError)
 	}
 	testModel = info.UpstreamModelName
 
-	apiType, _ := constant.ChannelType2APIType(channel.Type)
+	apiType, _ := common.ChannelType2APIType(channel.Type)
 	adaptor := relay.GetAdaptor(apiType)
 	if adaptor == nil {
-		return fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), nil
+		return fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), types.NewError(fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), types.ErrorCodeInvalidApiType)
 	}
 
 	request := buildTestRequest(testModel)
@@ -113,45 +117,45 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 
 	priceData, err := helper.ModelPriceHelper(c, info, 0, int(request.MaxTokens))
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeModelPriceError)
 	}
 
 	adaptor.Init(info)
 
 	convertedRequest, err := adaptor.ConvertOpenAIRequest(c, info, request)
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeConvertRequestFailed)
 	}
 	jsonData, err := json.Marshal(convertedRequest)
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeJsonMarshalFailed)
 	}
 	requestBody := bytes.NewBuffer(jsonData)
 	c.Request.Body = io.NopCloser(requestBody)
 	resp, err := adaptor.DoRequest(c, info, requestBody)
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeDoRequestFailed)
 	}
 	var httpResp *http.Response
 	if resp != nil {
 		httpResp = resp.(*http.Response)
 		if httpResp.StatusCode != http.StatusOK {
 			err := service.RelayErrorHandler(httpResp, true)
-			return fmt.Errorf("status code %d: %s", httpResp.StatusCode, err.Error.Message), err
+			return err, types.NewError(err, types.ErrorCodeBadResponse)
 		}
 	}
 	usageA, respErr := adaptor.DoResponse(c, httpResp, info)
 	if respErr != nil {
-		return fmt.Errorf("%s", respErr.Error.Message), respErr
+		return respErr, respErr
 	}
 	if usageA == nil {
-		return errors.New("usage is nil"), nil
+		return errors.New("usage is nil"), types.NewError(errors.New("usage is nil"), types.ErrorCodeBadResponseBody)
 	}
 	usage := usageA.(*dto.Usage)
 	result := w.Result()
 	respBody, err := io.ReadAll(result.Body)
 	if err != nil {
-		return err, nil
+		return err, types.NewError(err, types.ErrorCodeReadResponseBodyFailed)
 	}
 	info.PromptTokens = usage.PromptTokens
 
@@ -170,8 +174,19 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	consumedTime := float64(milliseconds) / 1000.0
 	other := service.GenerateTextOtherInfo(c, info, priceData.ModelRatio, priceData.GroupRatioInfo.GroupRatio, priceData.CompletionRatio,
 		usage.PromptTokensDetails.CachedTokens, priceData.CacheRatio, priceData.ModelPrice, priceData.GroupRatioInfo.GroupSpecialRatio)
-	model.RecordConsumeLog(c, 1, channel.Id, usage.PromptTokens, usage.CompletionTokens, info.OriginModelName, "模型测试",
-		quota, "模型测试", 0, quota, int(consumedTime), false, info.UsingGroup, other)
+	model.RecordConsumeLog(c, 1, model.RecordConsumeLogParams{
+		ChannelId:        channel.Id,
+		PromptTokens:     usage.PromptTokens,
+		CompletionTokens: usage.CompletionTokens,
+		ModelName:        info.OriginModelName,
+		TokenName:        "模型测试",
+		Quota:            quota,
+		Content:          "模型测试",
+		UseTimeSeconds:   int(consumedTime),
+		IsStream:         false,
+		Group:            info.UsingGroup,
+		Other:            other,
+	})
 	common.SysLog(fmt.Sprintf("testing channel #%d, response: \n%s", channel.Id, string(respBody)))
 	return nil, nil
 }
@@ -199,7 +214,7 @@ func buildTestRequest(model string) *dto.GeneralOpenAIRequest {
 			testRequest.MaxTokens = 50
 		}
 	} else if strings.Contains(model, "gemini") {
-		testRequest.MaxTokens = 300
+		testRequest.MaxTokens = 3000
 	} else {
 		testRequest.MaxTokens = 10
 	}
@@ -232,15 +247,15 @@ func TestChannel(c *gin.Context) {
 	}
 	testModel := c.Query("model")
 	tik := time.Now()
-	err, _ = testChannel(channel, testModel)
+	_, newAPIError := testChannel(channel, testModel)
 	tok := time.Now()
 	milliseconds := tok.Sub(tik).Milliseconds()
 	go channel.UpdateResponseTime(milliseconds)
 	consumedTime := float64(milliseconds) / 1000.0
-	if err != nil {
+	if newAPIError != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
-			"message": err.Error(),
+			"message": newAPIError.Error(),
 			"time":    consumedTime,
 		})
 		return
@@ -284,17 +299,15 @@ func testAllChannels(notify bool) error {
 		for _, channel := range channels {
 			isChannelEnabled := channel.Status == common.ChannelStatusEnabled
 			tik := time.Now()
-			err, openaiWithStatusErr := testChannel(channel, "")
+			err, newAPIError := testChannel(channel, "")
 			tok := time.Now()
 			milliseconds := tok.Sub(tik).Milliseconds()
 
 			shouldBanChannel := false
 
 			// request error disables the channel
-			if openaiWithStatusErr != nil {
-				oaiErr := openaiWithStatusErr.Error
-				err = errors.New(fmt.Sprintf("type %s, httpCode %d, code %v, message %s", oaiErr.Type, openaiWithStatusErr.StatusCode, oaiErr.Code, oaiErr.Message))
-				shouldBanChannel = service.ShouldDisableChannel(channel.Type, openaiWithStatusErr)
+			if err != nil {
+				shouldBanChannel = service.ShouldDisableChannel(channel.Type, newAPIError)
 			}
 
 			if milliseconds > disableThreshold {
@@ -308,7 +321,7 @@ func testAllChannels(notify bool) error {
 			}
 
 			// enable channel
-			if !isChannelEnabled && service.ShouldEnableChannel(err, openaiWithStatusErr, channel.Status) {
+			if !isChannelEnabled && service.ShouldEnableChannel(err, newAPIError, channel.Status) {
 				service.EnableChannel(channel.Id, channel.Name)
 			}
 
@@ -340,6 +353,10 @@ func TestAllChannels(c *gin.Context) {
 }
 
 func AutomaticallyTestChannels(frequency int) {
+	if frequency <= 0 {
+		common.SysLog("CHANNEL_TEST_FREQUENCY is not set or invalid, skipping automatic channel test")
+		return
+	}
 	for {
 		time.Sleep(time.Duration(frequency) * time.Minute)
 		common.SysLog("testing all channels")

controller/channel.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/model"
 	"strconv"
 	"strings"
@@ -125,7 +126,7 @@ func GetAllChannels(c *gin.Context) {
 			order = "id desc"
 		}
 
-		err := baseQuery.Order(order).Limit(pageSize).Offset((p-1)*pageSize).Omit("key").Find(&channelData).Error
+		err := baseQuery.Order(order).Limit(pageSize).Offset((p - 1) * pageSize).Omit("key").Find(&channelData).Error
 		if err != nil {
 			c.JSON(http.StatusOK, gin.H{"success": false, "message": err.Error()})
 			return
@@ -181,15 +182,15 @@ func FetchUpstreamModels(c *gin.Context) {
 		return
 	}
 
-	baseURL := common.ChannelBaseURLs[channel.Type]
+	baseURL := constant.ChannelBaseURLs[channel.Type]
 	if channel.GetBaseURL() != "" {
 		baseURL = channel.GetBaseURL()
 	}
 	url := fmt.Sprintf("%s/v1/models", baseURL)
 	switch channel.Type {
-	case common.ChannelTypeGemini:
+	case constant.ChannelTypeGemini:
 		url = fmt.Sprintf("%s/v1beta/openai/models", baseURL)
-	case common.ChannelTypeAli:
+	case constant.ChannelTypeAli:
 		url = fmt.Sprintf("%s/compatible-mode/v1/models", baseURL)
 	}
 	body, err := GetResponseBody("GET", url, channel, GetAuthHeader(channel.Key))
@@ -213,7 +214,7 @@ func FetchUpstreamModels(c *gin.Context) {
 	var ids []string
 	for _, model := range result.Data {
 		id := model.ID
-		if channel.Type == common.ChannelTypeGemini {
+		if channel.Type == constant.ChannelTypeGemini {
 			id = strings.TrimPrefix(id, "models/")
 		}
 		ids = append(ids, id)
@@ -227,7 +228,7 @@ func FetchUpstreamModels(c *gin.Context) {
 }
 
 func FixChannelsAbilities(c *gin.Context) {
-	count, err := model.FixAbility()
+	success, fails, err := model.FixAbility()
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -238,7 +239,10 @@ func FixChannelsAbilities(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data":    count,
+		"data": gin.H{
+			"success": success,
+			"fails":   fails,
+		},
 	})
 }
 
@@ -376,9 +380,47 @@ func GetChannel(c *gin.Context) {
 	return
 }
 
+type AddChannelRequest struct {
+	Mode         string                `json:"mode"`
+	MultiKeyMode constant.MultiKeyMode `json:"multi_key_mode"`
+	Channel      *model.Channel        `json:"channel"`
+}
+
+func getVertexArrayKeys(keys string) ([]string, error) {
+	if keys == "" {
+		return nil, nil
+	}
+	var keyArray []interface{}
+	err := common.Unmarshal([]byte(keys), &keyArray)
+	if err != nil {
+		return nil, fmt.Errorf("批量添加 Vertex AI 必须使用标准的JsonArray格式，例如[{key1}, {key2}...]，请检查输入: %w", err)
+	}
+	cleanKeys := make([]string, 0, len(keyArray))
+	for _, key := range keyArray {
+		var keyStr string
+		switch v := key.(type) {
+		case string:
+			keyStr = strings.TrimSpace(v)
+		default:
+			bytes, err := json.Marshal(v)
+			if err != nil {
+				return nil, fmt.Errorf("Vertex AI key JSON 编码失败: %w", err)
+			}
+			keyStr = string(bytes)
+		}
+		if keyStr != "" {
+			cleanKeys = append(cleanKeys, keyStr)
+		}
+	}
+	if len(cleanKeys) == 0 {
+		return nil, fmt.Errorf("批量添加 Vertex AI 的 keys 不能为空")
+	}
+	return cleanKeys, nil
+}
+
 func AddChannel(c *gin.Context) {
-	channel := model.Channel{}
-	err := c.ShouldBindJSON(&channel)
+	addChannelRequest := AddChannelRequest{}
+	err := c.ShouldBindJSON(&addChannelRequest)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -386,49 +428,120 @@ func AddChannel(c *gin.Context) {
 		})
 		return
 	}
-	channel.CreatedTime = common.GetTimestamp()
-	keys := strings.Split(channel.Key, "\n")
-	if channel.Type == common.ChannelTypeVertexAi {
-		if channel.Other == "" {
+
+	err = addChannelRequest.Channel.ValidateSettings()
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "channel setting 格式错误：" + err.Error(),
+		})
+		return
+	}
+
+	if addChannelRequest.Channel == nil || addChannelRequest.Channel.Key == "" {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "channel cannot be empty",
+		})
+		return
+	}
+
+	// Validate the length of the model name
+	for _, m := range addChannelRequest.Channel.GetModels() {
+		if len(m) > 255 {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": fmt.Sprintf("模型名称过长: %s", m),
+			})
+			return
+		}
+	}
+	if addChannelRequest.Channel.Type == constant.ChannelTypeVertexAi {
+		if addChannelRequest.Channel.Other == "" {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
 				"message": "部署地区不能为空",
 			})
 			return
 		} else {
-			if common.IsJsonStr(channel.Other) {
-				// must have default
-				regionMap := common.StrToMap(channel.Other)
-				if regionMap["default"] == nil {
-					c.JSON(http.StatusOK, gin.H{
-						"success": false,
-						"message": "部署地区必须包含default字段",
-					})
-					return
-				}
+			regionMap, err := common.StrToMap(addChannelRequest.Channel.Other)
+			if err != nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": "部署地区必须是标准的Json格式，例如{\"default\": \"us-central1\", \"region2\": \"us-east1\"}",
+				})
+				return
+			}
+			if regionMap["default"] == nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": "部署地区必须包含default字段",
+				})
+				return
 			}
 		}
-		keys = []string{channel.Key}
 	}
+
+	addChannelRequest.Channel.CreatedTime = common.GetTimestamp()
+	keys := make([]string, 0)
+	switch addChannelRequest.Mode {
+	case "multi_to_single":
+		addChannelRequest.Channel.ChannelInfo.IsMultiKey = true
+		addChannelRequest.Channel.ChannelInfo.MultiKeyMode = addChannelRequest.MultiKeyMode
+		if addChannelRequest.Channel.Type == constant.ChannelTypeVertexAi {
+			array, err := getVertexArrayKeys(addChannelRequest.Channel.Key)
+			if err != nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": err.Error(),
+				})
+				return
+			}
+			addChannelRequest.Channel.Key = strings.Join(array, "\n")
+		} else {
+			cleanKeys := make([]string, 0)
+			for _, key := range strings.Split(addChannelRequest.Channel.Key, "\n") {
+				if key == "" {
+					continue
+				}
+				key = strings.TrimSpace(key)
+				cleanKeys = append(cleanKeys, key)
+			}
+			addChannelRequest.Channel.Key = strings.Join(cleanKeys, "\n")
+		}
+		keys = []string{addChannelRequest.Channel.Key}
+	case "batch":
+		if addChannelRequest.Channel.Type == constant.ChannelTypeVertexAi {
+			// multi json
+			keys, err = getVertexArrayKeys(addChannelRequest.Channel.Key)
+			if err != nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": err.Error(),
+				})
+				return
+			}
+		} else {
+			keys = strings.Split(addChannelRequest.Channel.Key, "\n")
+		}
+	case "single":
+		keys = []string{addChannelRequest.Channel.Key}
+	default:
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "不支持的添加模式",
+		})
+		return
+	}
+
 	channels := make([]model.Channel, 0, len(keys))
 	for _, key := range keys {
 		if key == "" {
 			continue
 		}
-		localChannel := channel
+		localChannel := addChannelRequest.Channel
 		localChannel.Key = key
-		// Validate the length of the model name
-		models := strings.Split(localChannel.Models, ",")
-		for _, model := range models {
-			if len(model) > 255 {
-				c.JSON(http.StatusOK, gin.H{
-					"success": false,
-					"message": fmt.Sprintf("模型名称过长: %s", model),
-				})
-				return
-			}
-		}
-		channels = append(channels, localChannel)
+		channels = append(channels, *localChannel)
 	}
 	err = model.BatchInsertChannels(channels)
 	if err != nil {
@@ -613,7 +726,15 @@ func UpdateChannel(c *gin.Context) {
 		})
 		return
 	}
-	if channel.Type == common.ChannelTypeVertexAi {
+	err = channel.ValidateSettings()
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "channel setting 格式错误：" + err.Error(),
+		})
+		return
+	}
+	if channel.Type == constant.ChannelTypeVertexAi {
 		if channel.Other == "" {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
@@ -621,16 +742,20 @@ func UpdateChannel(c *gin.Context) {
 			})
 			return
 		} else {
-			if common.IsJsonStr(channel.Other) {
-				// must have default
-				regionMap := common.StrToMap(channel.Other)
-				if regionMap["default"] == nil {
-					c.JSON(http.StatusOK, gin.H{
-						"success": false,
-						"message": "部署地区必须包含default字段",
-					})
-					return
-				}
+			regionMap, err := common.StrToMap(channel.Other)
+			if err != nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": "部署地区必须是标准的Json格式，例如{\"default\": \"us-central1\", \"region2\": \"us-east1\"}",
+				})
+				return
+			}
+			if regionMap["default"] == nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": "部署地区必须包含default字段",
+				})
+				return
 			}
 		}
 	}
@@ -668,7 +793,7 @@ func FetchModels(c *gin.Context) {
 
 	baseURL := req.BaseURL
 	if baseURL == "" {
-		baseURL = common.ChannelBaseURLs[req.Type]
+		baseURL = constant.ChannelBaseURLs[req.Type]
 	}
 
 	client := &http.Client{}

controller/model.go

@@ -2,6 +2,7 @@ package controller
 
 import (
 	"fmt"
+	"github.com/gin-gonic/gin"
 	"github.com/samber/lo"
 	"net/http"
 	"one-api/common"
@@ -14,10 +15,7 @@ import (
 	"one-api/relay/channel/minimax"
 	"one-api/relay/channel/moonshot"
 	relaycommon "one-api/relay/common"
-	relayconstant "one-api/relay/constant"
 	"one-api/setting"
-
-	"github.com/gin-gonic/gin"
 )
 
 // https://platform.openai.com/docs/api-reference/models/list
@@ -26,30 +24,10 @@ var openAIModels []dto.OpenAIModels
 var openAIModelsMap map[string]dto.OpenAIModels
 var channelId2Models map[int][]string
 
-func getPermission() []dto.OpenAIModelPermission {
-	var permission []dto.OpenAIModelPermission
-	permission = append(permission, dto.OpenAIModelPermission{
-		Id:                 "modelperm-LwHkVFn8AcMItP432fKKDIKJ",
-		Object:             "model_permission",
-		Created:            1626777600,
-		AllowCreateEngine:  true,
-		AllowSampling:      true,
-		AllowLogprobs:      true,
-		AllowSearchIndices: false,
-		AllowView:          true,
-		AllowFineTuning:    false,
-		Organization:       "*",
-		Group:              nil,
-		IsBlocking:         false,
-	})
-	return permission
-}
-
 func init() {
 	// https://platform.openai.com/docs/models/model-endpoint-compatibility
-	permission := getPermission()
-	for i := 0; i < relayconstant.APITypeDummy; i++ {
-		if i == relayconstant.APITypeAIProxyLibrary {
+	for i := 0; i < constant.APITypeDummy; i++ {
+		if i == constant.APITypeAIProxyLibrary {
 			continue
 		}
 		adaptor := relay.GetAdaptor(i)
@@ -57,69 +35,51 @@ func init() {
 		modelNames := adaptor.GetModelList()
 		for _, modelName := range modelNames {
 			openAIModels = append(openAIModels, dto.OpenAIModels{
-				Id:         modelName,
-				Object:     "model",
-				Created:    1626777600,
-				OwnedBy:    channelName,
-				Permission: permission,
-				Root:       modelName,
-				Parent:     nil,
+				Id:      modelName,
+				Object:  "model",
+				Created: 1626777600,
+				OwnedBy: channelName,
 			})
 		}
 	}
 	for _, modelName := range ai360.ModelList {
 		openAIModels = append(openAIModels, dto.OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    ai360.ChannelName,
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
+			Id:      modelName,
+			Object:  "model",
+			Created: 1626777600,
+			OwnedBy: ai360.ChannelName,
 		})
 	}
 	for _, modelName := range moonshot.ModelList {
 		openAIModels = append(openAIModels, dto.OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    moonshot.ChannelName,
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
+			Id:      modelName,
+			Object:  "model",
+			Created: 1626777600,
+			OwnedBy: moonshot.ChannelName,
 		})
 	}
 	for _, modelName := range lingyiwanwu.ModelList {
 		openAIModels = append(openAIModels, dto.OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    lingyiwanwu.ChannelName,
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
+			Id:      modelName,
+			Object:  "model",
+			Created: 1626777600,
+			OwnedBy: lingyiwanwu.ChannelName,
 		})
 	}
 	for _, modelName := range minimax.ModelList {
 		openAIModels = append(openAIModels, dto.OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    minimax.ChannelName,
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
+			Id:      modelName,
+			Object:  "model",
+			Created: 1626777600,
+			OwnedBy: minimax.ChannelName,
 		})
 	}
 	for modelName, _ := range constant.MidjourneyModel2Action {
 		openAIModels = append(openAIModels, dto.OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "midjourney",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
+			Id:      modelName,
+			Object:  "model",
+			Created: 1626777600,
+			OwnedBy: "midjourney",
 		})
 	}
 	openAIModelsMap = make(map[string]dto.OpenAIModels)
@@ -127,9 +87,9 @@ func init() {
 		openAIModelsMap[aiModel.Id] = aiModel
 	}
 	channelId2Models = make(map[int][]string)
-	for i := 1; i <= common.ChannelTypeDummy; i++ {
-		apiType, success := relayconstant.ChannelType2APIType(i)
-		if !success || apiType == relayconstant.APITypeAIProxyLibrary {
+	for i := 1; i <= constant.ChannelTypeDummy; i++ {
+		apiType, success := common.ChannelType2APIType(i)
+		if !success || apiType == constant.APITypeAIProxyLibrary {
 			continue
 		}
 		meta := &relaycommon.RelayInfo{ChannelType: i}
@@ -144,11 +104,10 @@ func init() {
 
 func ListModels(c *gin.Context) {
 	userOpenAiModels := make([]dto.OpenAIModels, 0)
-	permission := getPermission()
 
-	modelLimitEnable := c.GetBool("token_model_limit_enabled")
+	modelLimitEnable := common.GetContextKeyBool(c, constant.ContextKeyTokenModelLimitEnabled)
 	if modelLimitEnable {
-		s, ok := c.Get("token_model_limit")
+		s, ok := common.GetContextKey(c, constant.ContextKeyTokenModelLimit)
 		var tokenModelLimit map[string]bool
 		if ok {
 			tokenModelLimit = s.(map[string]bool)
@@ -156,23 +115,22 @@ func ListModels(c *gin.Context) {
 			tokenModelLimit = map[string]bool{}
 		}
 		for allowModel, _ := range tokenModelLimit {
-			if _, ok := openAIModelsMap[allowModel]; ok {
-				userOpenAiModels = append(userOpenAiModels, openAIModelsMap[allowModel])
+			if oaiModel, ok := openAIModelsMap[allowModel]; ok {
+				oaiModel.SupportedEndpointTypes = model.GetModelSupportEndpointTypes(allowModel)
+				userOpenAiModels = append(userOpenAiModels, oaiModel)
 			} else {
 				userOpenAiModels = append(userOpenAiModels, dto.OpenAIModels{
-					Id:         allowModel,
-					Object:     "model",
-					Created:    1626777600,
-					OwnedBy:    "custom",
-					Permission: permission,
-					Root:       allowModel,
-					Parent:     nil,
+					Id:                     allowModel,
+					Object:                 "model",
+					Created:                1626777600,
+					OwnedBy:                "custom",
+					SupportedEndpointTypes: model.GetModelSupportEndpointTypes(allowModel),
 				})
 			}
 		}
 	} else {
 		userId := c.GetInt("id")
-		userGroup, err := model.GetUserGroup(userId, true)
+		userGroup, err := model.GetUserGroup(userId, false)
 		if err != nil {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
@@ -181,14 +139,14 @@ func ListModels(c *gin.Context) {
 			return
 		}
 		group := userGroup
-		tokenGroup := c.GetString("token_group")
+		tokenGroup := common.GetContextKeyString(c, constant.ContextKeyTokenGroup)
 		if tokenGroup != "" {
 			group = tokenGroup
 		}
 		var models []string
 		if tokenGroup == "auto" {
 			for _, autoGroup := range setting.AutoGroups {
-				groupModels := model.GetGroupModels(autoGroup)
+				groupModels := model.GetGroupEnabledModels(autoGroup)
 				for _, g := range groupModels {
 					if !common.StringsContains(models, g) {
 						models = append(models, g)
@@ -196,20 +154,19 @@ func ListModels(c *gin.Context) {
 				}
 			}
 		} else {
-			models = model.GetGroupModels(group)
+			models = model.GetGroupEnabledModels(group)
 		}
-		for _, s := range models {
-			if _, ok := openAIModelsMap[s]; ok {
-				userOpenAiModels = append(userOpenAiModels, openAIModelsMap[s])
+		for _, modelName := range models {
+			if oaiModel, ok := openAIModelsMap[modelName]; ok {
+				oaiModel.SupportedEndpointTypes = model.GetModelSupportEndpointTypes(modelName)
+				userOpenAiModels = append(userOpenAiModels, oaiModel)
 			} else {
 				userOpenAiModels = append(userOpenAiModels, dto.OpenAIModels{
-					Id:         s,
-					Object:     "model",
-					Created:    1626777600,
-					OwnedBy:    "custom",
-					Permission: permission,
-					Root:       s,
-					Parent:     nil,
+					Id:                     modelName,
+					Object:                 "model",
+					Created:                1626777600,
+					OwnedBy:                "custom",
+					SupportedEndpointTypes: model.GetModelSupportEndpointTypes(modelName),
 				})
 			}
 		}

controller/playground.go

@@ -3,45 +3,44 @@ package controller
 import (
 	"errors"
 	"fmt"
-	"net/http"
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
 	"one-api/middleware"
 	"one-api/model"
-	"one-api/service"
 	"one-api/setting"
+	"one-api/types"
 	"time"
 
 	"github.com/gin-gonic/gin"
 )
 
 func Playground(c *gin.Context) {
-	var openaiErr *dto.OpenAIErrorWithStatusCode
+	var newAPIError *types.NewAPIError
 
 	defer func() {
-		if openaiErr != nil {
-			c.JSON(openaiErr.StatusCode, gin.H{
-				"error": openaiErr.Error,
+		if newAPIError != nil {
+			c.JSON(newAPIError.StatusCode, gin.H{
+				"error": newAPIError.ToOpenAIError(),
 			})
 		}
 	}()
 
 	useAccessToken := c.GetBool("use_access_token")
 	if useAccessToken {
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("暂不支持使用 access token"), "access_token_not_supported", http.StatusBadRequest)
+		newAPIError = types.NewError(errors.New("暂不支持使用 access token"), types.ErrorCodeAccessDenied)
 		return
 	}
 
 	playgroundRequest := &dto.PlayGroundRequest{}
 	err := common.UnmarshalBodyReusable(c, playgroundRequest)
 	if err != nil {
-		openaiErr = service.OpenAIErrorWrapperLocal(err, "unmarshal_request_failed", http.StatusBadRequest)
+		newAPIError = types.NewError(err, types.ErrorCodeInvalidRequest)
 		return
 	}
 
 	if playgroundRequest.Model == "" {
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("请选择模型"), "model_required", http.StatusBadRequest)
+		newAPIError = types.NewError(errors.New("请选择模型"), types.ErrorCodeInvalidRequest)
 		return
 	}
 	c.Set("original_model", playgroundRequest.Model)
@@ -52,26 +51,32 @@ func Playground(c *gin.Context) {
 		group = userGroup
 	} else {
 		if !setting.GroupInUserUsableGroups(group) && group != userGroup {
-			openaiErr = service.OpenAIErrorWrapperLocal(errors.New("无权访问该分组"), "group_not_allowed", http.StatusForbidden)
+			newAPIError = types.NewError(errors.New("无权访问该分组"), types.ErrorCodeAccessDenied)
 			return
 		}
 		c.Set("group", group)
 	}
-	c.Set("token_name", "playground-"+group)
-	channel, finalGroup, err := model.CacheGetRandomSatisfiedChannel(c, group, playgroundRequest.Model, 0)
+
+	userId := c.GetInt("id")
+	//c.Set("token_name", "playground-"+group)
+	tempToken := &model.Token{
+		UserId: userId,
+		Name:   fmt.Sprintf("playground-%s", group),
+		Group:  group,
+	}
+	_ = middleware.SetupContextForToken(c, tempToken)
+	_, err = getChannel(c, group, playgroundRequest.Model, 0)
 	if err != nil {
-		message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", finalGroup, playgroundRequest.Model)
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New(message), "get_playground_channel_failed", http.StatusInternalServerError)
+		newAPIError = types.NewError(err, types.ErrorCodeGetChannelFailed)
 		return
 	}
-	middleware.SetupContextForSelectedChannel(c, channel, playgroundRequest.Model)
-	c.Set(constant.ContextKeyRequestStartTime, time.Now())
+	//middleware.SetupContextForSelectedChannel(c, channel, playgroundRequest.Model)
+	common.SetContextKey(c, constant.ContextKeyRequestStartTime, time.Now())
 
 	// Write user context to ensure acceptUnsetRatio is available
-	userId := c.GetInt("id")
 	userCache, err := model.GetUserCache(userId)
 	if err != nil {
-		openaiErr = service.OpenAIErrorWrapperLocal(err, "get_user_cache_failed", http.StatusInternalServerError)
+		newAPIError = types.NewError(err, types.ErrorCodeQueryDataError)
 		return
 	}
 	userCache.WriteContext(c)

controller/relay.go

@@ -8,23 +8,24 @@ import (
 	"log"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	constant2 "one-api/constant"
 	"one-api/dto"
 	"one-api/middleware"
 	"one-api/model"
 	"one-api/relay"
-	"one-api/relay/constant"
 	relayconstant "one-api/relay/constant"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
 	"github.com/gorilla/websocket"
 )
 
-func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
-	var err *dto.OpenAIErrorWithStatusCode
+func relayHandler(c *gin.Context, relayMode int) *types.NewAPIError {
+	var err *types.NewAPIError
 	switch relayMode {
 	case relayconstant.RelayModeImagesGenerations, relayconstant.RelayModeImagesEdits:
 		err = relay.ImageHelper(c)
@@ -55,43 +56,43 @@ func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode
 		userGroup := c.GetString("group")
 		channelId := c.GetInt("channel_id")
 		other := make(map[string]interface{})
-		other["error_type"] = err.Error.Type
-		other["error_code"] = err.Error.Code
+		other["error_type"] = err.ErrorType
+		other["error_code"] = err.GetErrorCode()
 		other["status_code"] = err.StatusCode
 		other["channel_id"] = channelId
 		other["channel_name"] = c.GetString("channel_name")
 		other["channel_type"] = c.GetInt("channel_type")
 
-		model.RecordErrorLog(c, userId, channelId, modelName, tokenName, err.Error.Message, tokenId, 0, false, userGroup, other)
+		model.RecordErrorLog(c, userId, channelId, modelName, tokenName, err.Error(), tokenId, 0, false, userGroup, other)
 	}
 
 	return err
 }
 
 func Relay(c *gin.Context) {
-	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
+	relayMode := relayconstant.Path2RelayMode(c.Request.URL.Path)
 	requestId := c.GetString(common.RequestIdKey)
 	group := c.GetString("group")
 	originalModel := c.GetString("original_model")
-	var openaiErr *dto.OpenAIErrorWithStatusCode
+	var newAPIError *types.NewAPIError
 
 	for i := 0; i <= common.RetryTimes; i++ {
 		channel, err := getChannel(c, group, originalModel, i)
 		if err != nil {
 			common.LogError(c, err.Error())
-			openaiErr = service.OpenAIErrorWrapperLocal(err, "get_channel_failed", http.StatusInternalServerError)
+			newAPIError = types.NewError(err, types.ErrorCodeGetChannelFailed)
 			break
 		}
 
-		openaiErr = relayRequest(c, relayMode, channel)
+		newAPIError = relayRequest(c, relayMode, channel)
 
-		if openaiErr == nil {
+		if newAPIError == nil {
 			return // 成功处理请求，直接返回
 		}
 
-		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), openaiErr)
+		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), newAPIError)
 
-		if !shouldRetry(c, openaiErr, common.RetryTimes-i) {
+		if !shouldRetry(c, newAPIError, common.RetryTimes-i) {
 			break
 		}
 	}
@@ -101,14 +102,14 @@ func Relay(c *gin.Context) {
 		common.LogInfo(c, retryLogStr)
 	}
 
-	if openaiErr != nil {
-		if openaiErr.StatusCode == http.StatusTooManyRequests {
-			common.LogError(c, fmt.Sprintf("origin 429 error: %s", openaiErr.Error.Message))
-			openaiErr.Error.Message = "当前分组上游负载已饱和，请稍后再试"
+	if newAPIError != nil {
+		if newAPIError.StatusCode == http.StatusTooManyRequests {
+			common.LogError(c, fmt.Sprintf("origin 429 error: %s", newAPIError.Error()))
+			newAPIError.SetMessage("当前分组上游负载已饱和，请稍后再试")
 		}
-		openaiErr.Error.Message = common.MessageWithRequestId(openaiErr.Error.Message, requestId)
-		c.JSON(openaiErr.StatusCode, gin.H{
-			"error": openaiErr.Error,
+		newAPIError.SetMessage(common.MessageWithRequestId(newAPIError.Error(), requestId))
+		c.JSON(newAPIError.StatusCode, gin.H{
+			"error": newAPIError.ToOpenAIError(),
 		})
 	}
 }
@@ -127,35 +128,34 @@ func WssRelay(c *gin.Context) {
 	defer ws.Close()
 
 	if err != nil {
-		openaiErr := service.OpenAIErrorWrapper(err, "get_channel_failed", http.StatusInternalServerError)
-		helper.WssError(c, ws, openaiErr.Error)
+		helper.WssError(c, ws, types.NewError(err, types.ErrorCodeGetChannelFailed).ToOpenAIError())
 		return
 	}
 
-	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
+	relayMode := relayconstant.Path2RelayMode(c.Request.URL.Path)
 	requestId := c.GetString(common.RequestIdKey)
 	group := c.GetString("group")
 	//wss://api.openai.com/v1/realtime?model=gpt-4o-realtime-preview-2024-10-01
 	originalModel := c.GetString("original_model")
-	var openaiErr *dto.OpenAIErrorWithStatusCode
+	var newAPIError *types.NewAPIError
 
 	for i := 0; i <= common.RetryTimes; i++ {
 		channel, err := getChannel(c, group, originalModel, i)
 		if err != nil {
 			common.LogError(c, err.Error())
-			openaiErr = service.OpenAIErrorWrapperLocal(err, "get_channel_failed", http.StatusInternalServerError)
+			newAPIError = types.NewError(err, types.ErrorCodeGetChannelFailed)
 			break
 		}
 
-		openaiErr = wssRequest(c, ws, relayMode, channel)
+		newAPIError = wssRequest(c, ws, relayMode, channel)
 
-		if openaiErr == nil {
+		if newAPIError == nil {
 			return // 成功处理请求，直接返回
 		}
 
-		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), openaiErr)
+		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), newAPIError)
 
-		if !shouldRetry(c, openaiErr, common.RetryTimes-i) {
+		if !shouldRetry(c, newAPIError, common.RetryTimes-i) {
 			break
 		}
 	}
@@ -165,12 +165,12 @@ func WssRelay(c *gin.Context) {
 		common.LogInfo(c, retryLogStr)
 	}
 
-	if openaiErr != nil {
-		if openaiErr.StatusCode == http.StatusTooManyRequests {
-			openaiErr.Error.Message = "当前分组上游负载已饱和，请稍后再试"
+	if newAPIError != nil {
+		if newAPIError.StatusCode == http.StatusTooManyRequests {
+			newAPIError.SetMessage("当前分组上游负载已饱和，请稍后再试")
 		}
-		openaiErr.Error.Message = common.MessageWithRequestId(openaiErr.Error.Message, requestId)
-		helper.WssError(c, ws, openaiErr.Error)
+		newAPIError.SetMessage(common.MessageWithRequestId(newAPIError.Error(), requestId))
+		helper.WssError(c, ws, newAPIError.ToOpenAIError())
 	}
 }
 
@@ -179,27 +179,25 @@ func RelayClaude(c *gin.Context) {
 	requestId := c.GetString(common.RequestIdKey)
 	group := c.GetString("group")
 	originalModel := c.GetString("original_model")
-	var claudeErr *dto.ClaudeErrorWithStatusCode
+	var newAPIError *types.NewAPIError
 
 	for i := 0; i <= common.RetryTimes; i++ {
 		channel, err := getChannel(c, group, originalModel, i)
 		if err != nil {
 			common.LogError(c, err.Error())
-			claudeErr = service.ClaudeErrorWrapperLocal(err, "get_channel_failed", http.StatusInternalServerError)
+			newAPIError = types.NewError(err, types.ErrorCodeGetChannelFailed)
 			break
 		}
 
-		claudeErr = claudeRequest(c, channel)
+		newAPIError = claudeRequest(c, channel)
 
-		if claudeErr == nil {
+		if newAPIError == nil {
 			return // 成功处理请求，直接返回
 		}
 
-		openaiErr := service.ClaudeErrorToOpenAIError(claudeErr)
+		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), newAPIError)
 
-		go processChannelError(c, channel.Id, channel.Type, channel.Name, channel.GetAutoBan(), openaiErr)
-
-		if !shouldRetry(c, openaiErr, common.RetryTimes-i) {
+		if !shouldRetry(c, newAPIError, common.RetryTimes-i) {
 			break
 		}
 	}
@@ -209,30 +207,30 @@ func RelayClaude(c *gin.Context) {
 		common.LogInfo(c, retryLogStr)
 	}
 
-	if claudeErr != nil {
-		claudeErr.Error.Message = common.MessageWithRequestId(claudeErr.Error.Message, requestId)
-		c.JSON(claudeErr.StatusCode, gin.H{
+	if newAPIError != nil {
+		newAPIError.SetMessage(common.MessageWithRequestId(newAPIError.Error(), requestId))
+		c.JSON(newAPIError.StatusCode, gin.H{
 			"type":  "error",
-			"error": claudeErr.Error,
+			"error": newAPIError.ToClaudeError(),
 		})
 	}
 }
 
-func relayRequest(c *gin.Context, relayMode int, channel *model.Channel) *dto.OpenAIErrorWithStatusCode {
+func relayRequest(c *gin.Context, relayMode int, channel *model.Channel) *types.NewAPIError {
 	addUsedChannel(c, channel.Id)
 	requestBody, _ := common.GetRequestBody(c)
 	c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
 	return relayHandler(c, relayMode)
 }
 
-func wssRequest(c *gin.Context, ws *websocket.Conn, relayMode int, channel *model.Channel) *dto.OpenAIErrorWithStatusCode {
+func wssRequest(c *gin.Context, ws *websocket.Conn, relayMode int, channel *model.Channel) *types.NewAPIError {
 	addUsedChannel(c, channel.Id)
 	requestBody, _ := common.GetRequestBody(c)
 	c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
 	return relay.WssHelper(c, ws)
 }
 
-func claudeRequest(c *gin.Context, channel *model.Channel) *dto.ClaudeErrorWithStatusCode {
+func claudeRequest(c *gin.Context, channel *model.Channel) *types.NewAPIError {
 	addUsedChannel(c, channel.Id)
 	requestBody, _ := common.GetRequestBody(c)
 	c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
@@ -259,19 +257,25 @@ func getChannel(c *gin.Context, group, originalModel string, retryCount int) (*m
 			AutoBan: &autoBanInt,
 		}, nil
 	}
-	channel, _, err := model.CacheGetRandomSatisfiedChannel(c, group, originalModel, retryCount)
+	channel, selectGroup, err := model.CacheGetRandomSatisfiedChannel(c, group, originalModel, retryCount)
 	if err != nil {
-		return nil, errors.New(fmt.Sprintf("获取重试渠道失败: %s", err.Error()))
+		if group == "auto" {
+			return nil, errors.New(fmt.Sprintf("获取自动分组下模型 %s 的可用渠道失败: %s", originalModel, err.Error()))
+		}
+		return nil, errors.New(fmt.Sprintf("获取分组 %s 下模型 %s 的可用渠道失败: %s", selectGroup, originalModel, err.Error()))
 	}
 	middleware.SetupContextForSelectedChannel(c, channel, originalModel)
 	return channel, nil
 }
 
-func shouldRetry(c *gin.Context, openaiErr *dto.OpenAIErrorWithStatusCode, retryTimes int) bool {
+func shouldRetry(c *gin.Context, openaiErr *types.NewAPIError, retryTimes int) bool {
 	if openaiErr == nil {
 		return false
 	}
-	if openaiErr.LocalError {
+	if types.IsChannelError(openaiErr) {
+		return true
+	}
+	if types.IsLocalError(openaiErr) {
 		return false
 	}
 	if retryTimes <= 0 {
@@ -295,7 +299,7 @@ func shouldRetry(c *gin.Context, openaiErr *dto.OpenAIErrorWithStatusCode, retry
 	}
 	if openaiErr.StatusCode == http.StatusBadRequest {
 		channelType := c.GetInt("channel_type")
-		if channelType == common.ChannelTypeAnthropic {
+		if channelType == constant.ChannelTypeAnthropic {
 			return true
 		}
 		return false
@@ -310,12 +314,12 @@ func shouldRetry(c *gin.Context, openaiErr *dto.OpenAIErrorWithStatusCode, retry
 	return true
 }
 
-func processChannelError(c *gin.Context, channelId int, channelType int, channelName string, autoBan bool, err *dto.OpenAIErrorWithStatusCode) {
+func processChannelError(c *gin.Context, channelId int, channelType int, channelName string, autoBan bool, err *types.NewAPIError) {
 	// 不要使用context获取渠道信息，异步处理时可能会出现渠道信息不一致的情况
 	// do not use context to get channel info, there may be inconsistent channel info when processing asynchronously
-	common.LogError(c, fmt.Sprintf("relay error (channel #%d, status code: %d): %s", channelId, err.StatusCode, err.Error.Message))
+	common.LogError(c, fmt.Sprintf("relay error (channel #%d, status code: %d): %s", channelId, err.StatusCode, err.Error()))
 	if service.ShouldDisableChannel(channelType, err) && autoBan {
-		service.DisableChannel(channelId, channelName, err.Error.Message)
+		service.DisableChannel(channelId, channelName, err.Error())
 	}
 }
 
@@ -388,9 +392,10 @@ func RelayTask(c *gin.Context) {
 		retryTimes = 0
 	}
 	for i := 0; shouldRetryTaskRelay(c, channelId, taskErr, retryTimes) && i < retryTimes; i++ {
-		channel, _, err := model.CacheGetRandomSatisfiedChannel(c, group, originalModel, i)
+		channel, err := getChannel(c, group, originalModel, i)
 		if err != nil {
 			common.LogError(c, fmt.Sprintf("CacheGetRandomSatisfiedChannel failed: %s", err.Error()))
+			taskErr = service.TaskErrorWrapperLocal(err, "get_channel_failed", http.StatusInternalServerError)
 			break
 		}
 		channelId = channel.Id
@@ -398,7 +403,7 @@ func RelayTask(c *gin.Context) {
 		useChannel = append(useChannel, fmt.Sprintf("%d", channelId))
 		c.Set("use_channel", useChannel)
 		common.LogInfo(c, fmt.Sprintf("using channel #%d to retry (remain times %d)", channel.Id, i))
-		middleware.SetupContextForSelectedChannel(c, channel, originalModel)
+		//middleware.SetupContextForSelectedChannel(c, channel, originalModel)
 
 		requestBody, err := common.GetRequestBody(c)
 		c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))

controller/task.go

@@ -74,8 +74,8 @@ func UpdateTaskByPlatform(platform constant.TaskPlatform, taskChannelM map[int][
 		//_ = UpdateMidjourneyTaskAll(context.Background(), tasks)
 	case constant.TaskPlatformSuno:
 		_ = UpdateSunoTaskAll(context.Background(), taskChannelM, taskM)
-	case constant.TaskPlatformKling:
-		_ = UpdateVideoTaskAll(context.Background(), taskChannelM, taskM)
+	case constant.TaskPlatformKling, constant.TaskPlatformJimeng:
+		_ = UpdateVideoTaskAll(context.Background(), platform, taskChannelM, taskM)
 	default:
 		common.SysLog("未知平台")
 	}

controller/task_video.go

@@ -2,27 +2,26 @@ package controller
 
 import (
 	"context"
-	"encoding/json"
 	"fmt"
 	"io"
-	"net/http"
 	"one-api/common"
 	"one-api/constant"
 	"one-api/model"
 	"one-api/relay"
 	"one-api/relay/channel"
+	"time"
 )
 
-func UpdateVideoTaskAll(ctx context.Context, taskChannelM map[int][]string, taskM map[string]*model.Task) error {
+func UpdateVideoTaskAll(ctx context.Context, platform constant.TaskPlatform, taskChannelM map[int][]string, taskM map[string]*model.Task) error {
 	for channelId, taskIds := range taskChannelM {
-		if err := updateVideoTaskAll(ctx, channelId, taskIds, taskM); err != nil {
+		if err := updateVideoTaskAll(ctx, platform, channelId, taskIds, taskM); err != nil {
 			common.LogError(ctx, fmt.Sprintf("Channel #%d failed to update video async tasks: %s", channelId, err.Error()))
 		}
 	}
 	return nil
 }
 
-func updateVideoTaskAll(ctx context.Context, channelId int, taskIds []string, taskM map[string]*model.Task) error {
+func updateVideoTaskAll(ctx context.Context, platform constant.TaskPlatform, channelId int, taskIds []string, taskM map[string]*model.Task) error {
 	common.LogInfo(ctx, fmt.Sprintf("Channel #%d pending video tasks: %d", channelId, len(taskIds)))
 	if len(taskIds) == 0 {
 		return nil
@@ -39,7 +38,7 @@ func updateVideoTaskAll(ctx context.Context, channelId int, taskIds []string, ta
 		}
 		return fmt.Errorf("CacheGetChannel failed: %w", err)
 	}
-	adaptor := relay.GetTaskAdaptor(constant.TaskPlatformKling)
+	adaptor := relay.GetTaskAdaptor(platform)
 	if adaptor == nil {
 		return fmt.Errorf("video adaptor not found")
 	}
@@ -52,74 +51,68 @@ func updateVideoTaskAll(ctx context.Context, channelId int, taskIds []string, ta
 }
 
 func updateVideoSingleTask(ctx context.Context, adaptor channel.TaskAdaptor, channel *model.Channel, taskId string, taskM map[string]*model.Task) error {
-	baseURL := common.ChannelBaseURLs[channel.Type]
+	baseURL := constant.ChannelBaseURLs[channel.Type]
 	if channel.GetBaseURL() != "" {
 		baseURL = channel.GetBaseURL()
 	}
-	resp, err := adaptor.FetchTask(baseURL, channel.Key, map[string]any{
-		"task_id": taskId,
-	})
-	if err != nil {
-		return fmt.Errorf("FetchTask failed for task %s: %w", taskId, err)
-	}
-	if resp.StatusCode != http.StatusOK {
-		return fmt.Errorf("Get Video Task status code: %d", resp.StatusCode)
-	}
-	defer resp.Body.Close()
-	responseBody, err := io.ReadAll(resp.Body)
-	if err != nil {
-		return fmt.Errorf("ReadAll failed for task %s: %w", taskId, err)
-	}
-
-	var responseItem map[string]interface{}
-	err = json.Unmarshal(responseBody, &responseItem)
-	if err != nil {
-		common.LogError(ctx, fmt.Sprintf("Failed to parse video task response body: %v, body: %s", err, string(responseBody)))
-		return fmt.Errorf("Unmarshal failed for task %s: %w", taskId, err)
-	}
-
-	code, _ := responseItem["code"].(float64)
-	if code != 0 {
-		return fmt.Errorf("video task fetch failed for task %s", taskId)
-	}
-
-	data, ok := responseItem["data"].(map[string]interface{})
-	if !ok {
-		common.LogError(ctx, fmt.Sprintf("Video task data format error: %s", string(responseBody)))
-		return fmt.Errorf("video task data format error for task %s", taskId)
-	}
 
 	task := taskM[taskId]
 	if task == nil {
 		common.LogError(ctx, fmt.Sprintf("Task %s not found in taskM", taskId))
 		return fmt.Errorf("task %s not found", taskId)
 	}
-
-	if status, ok := data["task_status"].(string); ok {
-		switch status {
-		case "submitted", "queued":
-			task.Status = model.TaskStatusSubmitted
-		case "processing":
-			task.Status = model.TaskStatusInProgress
-		case "succeed":
-			task.Status = model.TaskStatusSuccess
-			task.Progress = "100%"
-			if url, err := adaptor.ParseResultUrl(responseItem); err == nil {
-				task.FailReason = url
-			} else {
-				common.LogWarn(ctx, fmt.Sprintf("Failed to get url from body for task %s: %s", task.TaskID, err.Error()))
-			}
-		case "failed":
-			task.Status = model.TaskStatusFailure
-			task.Progress = "100%"
-			if reason, ok := data["fail_reason"].(string); ok {
-				task.FailReason = reason
-			}
-		}
+	resp, err := adaptor.FetchTask(baseURL, channel.Key, map[string]any{
+		"task_id": taskId,
+		"action":  task.Action,
+	})
+	if err != nil {
+		return fmt.Errorf("fetchTask failed for task %s: %w", taskId, err)
+	}
+	//if resp.StatusCode != http.StatusOK {
+	//return fmt.Errorf("get Video Task status code: %d", resp.StatusCode)
+	//}
+	defer resp.Body.Close()
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return fmt.Errorf("readAll failed for task %s: %w", taskId, err)
 	}
 
-	// If task failed, refund quota
-	if task.Status == model.TaskStatusFailure {
+	taskResult, err := adaptor.ParseTaskResult(responseBody)
+	if err != nil {
+		return fmt.Errorf("parseTaskResult failed for task %s: %w", taskId, err)
+	}
+	//if taskResult.Code != 0 {
+	//	return fmt.Errorf("video task fetch failed for task %s", taskId)
+	//}
+
+	now := time.Now().Unix()
+	if taskResult.Status == "" {
+		return fmt.Errorf("task %s status is empty", taskId)
+	}
+	task.Status = model.TaskStatus(taskResult.Status)
+	switch taskResult.Status {
+	case model.TaskStatusSubmitted:
+		task.Progress = "10%"
+	case model.TaskStatusQueued:
+		task.Progress = "20%"
+	case model.TaskStatusInProgress:
+		task.Progress = "30%"
+		if task.StartTime == 0 {
+			task.StartTime = now
+		}
+	case model.TaskStatusSuccess:
+		task.Progress = "100%"
+		if task.FinishTime == 0 {
+			task.FinishTime = now
+		}
+		task.FailReason = taskResult.Url
+	case model.TaskStatusFailure:
+		task.Status = model.TaskStatusFailure
+		task.Progress = "100%"
+		if task.FinishTime == 0 {
+			task.FinishTime = now
+		}
+		task.FailReason = taskResult.Reason
 		common.LogInfo(ctx, fmt.Sprintf("Task %s failed: %s", task.TaskID, task.FailReason))
 		quota := task.Quota
 		if quota != 0 {
@@ -129,6 +122,11 @@ func updateVideoSingleTask(ctx context.Context, adaptor channel.TaskAdaptor, cha
 			logContent := fmt.Sprintf("Video async task failed %s, refund %s", task.TaskID, common.LogQuota(quota))
 			model.RecordLog(task.UserId, model.LogTypeSystem, logContent)
 		}
+	default:
+		return fmt.Errorf("unknown task status %s for task %s", taskResult.Status, taskId)
+	}
+	if taskResult.Progress != "" {
+		task.Progress = taskResult.Progress
 	}
 
 	task.Data = responseBody

controller/user.go

@@ -6,6 +6,7 @@ import (
 	"net/http"
 	"net/url"
 	"one-api/common"
+	"one-api/dto"
 	"one-api/model"
 	"one-api/setting"
 	"strconv"
@@ -246,15 +247,15 @@ func Register(c *gin.Context) {
 }
 
 func GetAllUsers(c *gin.Context) {
-	p, _ := strconv.Atoi(c.Query("p"))
-	pageSize, _ := strconv.Atoi(c.Query("page_size"))
-	if p < 1 {
-		p = 1
+	pageInfo, err := common.GetPageQuery(c)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "parse page query failed",
+		})
+		return
 	}
-	if pageSize < 0 {
-		pageSize = common.ItemsPerPage
-	}
-	users, total, err := model.GetAllUsers((p-1)*pageSize, pageSize)
+	users, total, err := model.GetAllUsers(pageInfo)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -262,15 +263,13 @@ func GetAllUsers(c *gin.Context) {
 		})
 		return
 	}
+
+	pageInfo.SetTotal(int(total))
+	pageInfo.SetItems(users)
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data": gin.H{
-			"items":     users,
-			"total":     total,
-			"page":      p,
-			"page_size": pageSize,
-		},
+		"data":    pageInfo,
 	})
 	return
 }
@@ -489,7 +488,7 @@ func GetUserModels(c *gin.Context) {
 	groups := setting.GetUserUsableGroups(user.Group)
 	var models []string
 	for group := range groups {
-		for _, g := range model.GetGroupModels(group) {
+		for _, g := range model.GetGroupEnabledModels(group) {
 			if !common.StringsContains(models, g) {
 				models = append(models, g)
 			}
@@ -963,7 +962,7 @@ func UpdateUserSetting(c *gin.Context) {
 	}
 
 	// 验证预警类型
-	if req.QuotaWarningType != constant.NotifyTypeEmail && req.QuotaWarningType != constant.NotifyTypeWebhook {
+	if req.QuotaWarningType != dto.NotifyTypeEmail && req.QuotaWarningType != dto.NotifyTypeWebhook {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
 			"message": "无效的预警类型",
@@ -981,7 +980,7 @@ func UpdateUserSetting(c *gin.Context) {
 	}
 
 	// 如果是webhook类型,验证webhook地址
-	if req.QuotaWarningType == constant.NotifyTypeWebhook {
+	if req.QuotaWarningType == dto.NotifyTypeWebhook {
 		if req.WebhookUrl == "" {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
@@ -1000,7 +999,7 @@ func UpdateUserSetting(c *gin.Context) {
 	}
 
 	// 如果是邮件类型，验证邮箱地址
-	if req.QuotaWarningType == constant.NotifyTypeEmail && req.NotificationEmail != "" {
+	if req.QuotaWarningType == dto.NotifyTypeEmail && req.NotificationEmail != "" {
 		// 验证邮箱格式
 		if !strings.Contains(req.NotificationEmail, "@") {
 			c.JSON(http.StatusOK, gin.H{
@@ -1022,24 +1021,24 @@ func UpdateUserSetting(c *gin.Context) {
 	}
 
 	// 构建设置
-	settings := map[string]interface{}{
-		constant.UserSettingNotifyType:            req.QuotaWarningType,
-		constant.UserSettingQuotaWarningThreshold: req.QuotaWarningThreshold,
-		"accept_unset_model_ratio_model":          req.AcceptUnsetModelRatioModel,
-		constant.UserSettingRecordIpLog:           req.RecordIpLog,
+	settings := dto.UserSetting{
+		NotifyType:            req.QuotaWarningType,
+		QuotaWarningThreshold: req.QuotaWarningThreshold,
+		AcceptUnsetRatioModel: req.AcceptUnsetModelRatioModel,
+		RecordIpLog:           req.RecordIpLog,
 	}
 
 	// 如果是webhook类型,添加webhook相关设置
-	if req.QuotaWarningType == constant.NotifyTypeWebhook {
-		settings[constant.UserSettingWebhookUrl] = req.WebhookUrl
+	if req.QuotaWarningType == dto.NotifyTypeWebhook {
+		settings.WebhookUrl = req.WebhookUrl
 		if req.WebhookSecret != "" {
-			settings[constant.UserSettingWebhookSecret] = req.WebhookSecret
+			settings.WebhookSecret = req.WebhookSecret
 		}
 	}
 
 	// 如果提供了通知邮箱，添加到设置中
-	if req.QuotaWarningType == constant.NotifyTypeEmail && req.NotificationEmail != "" {
-		settings[constant.UserSettingNotificationEmail] = req.NotificationEmail
+	if req.QuotaWarningType == dto.NotifyTypeEmail && req.NotificationEmail != "" {
+		settings.NotificationEmail = req.NotificationEmail
 	}
 
 	// 更新用户设置

docker-compose.yml

@@ -16,7 +16,7 @@ services:
       - REDIS_CONN_STRING=redis://redis
       - TZ=Asia/Shanghai
       - ERROR_LOG_ENABLED=true # 是否启用错误日志记录
-    #      - TIKTOKEN_CACHE_DIR=./tiktoken_cache  # 如果需要使用tiktoken_cache，请取消注释
+    #      - STREAMING_TIMEOUT=120  # 流模式无响应超时时间，单位秒，默认120秒，如果出现空补全可以尝试改为更大值
     #      - SESSION_SECRET=random_string  # 多机部署时设置，必须修改这个随机字符串！！！！！！！
     #      - NODE_TYPE=slave  # Uncomment for slave node in multi-node deployment
     #      - SYNC_FREQUENCY=60  # Uncomment if regular database syncing is needed

dto/channel_settings.go

@@ -0,0 +1,7 @@
+package dto
+
+type ChannelSettings struct {
+	ForceFormat       bool   `json:"force_format,omitempty"`
+	ThinkingToContent bool   `json:"thinking_to_content,omitempty"`
+	Proxy             string `json:"proxy"`
+}

dto/claude.go

@@ -3,6 +3,7 @@ package dto
 import (
 	"encoding/json"
 	"one-api/common"
+	"one-api/types"
 )
 
 type ClaudeMetadata struct {
@@ -228,7 +229,7 @@ type ClaudeResponse struct {
 	Completion   string               `json:"completion,omitempty"`
 	StopReason   string               `json:"stop_reason,omitempty"`
 	Model        string               `json:"model,omitempty"`
-	Error        *ClaudeError         `json:"error,omitempty"`
+	Error        *types.ClaudeError   `json:"error,omitempty"`
 	Usage        *ClaudeUsage         `json:"usage,omitempty"`
 	Index        *int                 `json:"index,omitempty"`
 	ContentBlock *ClaudeMediaMessage  `json:"content_block,omitempty"`

dto/error.go

@@ -1,5 +1,7 @@
 package dto
 
+import "one-api/types"
+
 type OpenAIError struct {
 	Message string `json:"message"`
 	Type    string `json:"type"`
@@ -14,11 +16,11 @@ type OpenAIErrorWithStatusCode struct {
 }
 
 type GeneralErrorResponse struct {
-	Error    OpenAIError `json:"error"`
-	Message  string      `json:"message"`
-	Msg      string      `json:"msg"`
-	Err      string      `json:"err"`
-	ErrorMsg string      `json:"error_msg"`
+	Error    types.OpenAIError `json:"error"`
+	Message  string            `json:"message"`
+	Msg      string            `json:"msg"`
+	Err      string            `json:"err"`
+	ErrorMsg string            `json:"error_msg"`
 	Header   struct {
 		Message string `json:"message"`
 	} `json:"header"`

dto/midjourney.go

@@ -57,6 +57,8 @@ type MidjourneyDto struct {
 	StartTime   int64       `json:"startTime"`
 	FinishTime  int64       `json:"finishTime"`
 	ImageUrl    string      `json:"imageUrl"`
+	VideoUrl    string      `json:"videoUrl"`
+	VideoUrls   []ImgUrls   `json:"videoUrls"`
 	Status      string      `json:"status"`
 	Progress    string      `json:"progress"`
 	FailReason  string      `json:"failReason"`
@@ -65,6 +67,10 @@ type MidjourneyDto struct {
 	Properties  *Properties `json:"properties"`
 }
 
+type ImgUrls struct {
+	Url string `json:"url"`
+}
+
 type MidjourneyStatus struct {
 	Status int `json:"status"`
 }

dto/openai_request.go

@@ -65,8 +65,8 @@ type GeneralOpenAIRequest struct {
 
 func (r *GeneralOpenAIRequest) ToMap() map[string]any {
 	result := make(map[string]any)
-	data, _ := common.EncodeJson(r)
-	_ = common.DecodeJson(data, &result)
+	data, _ := common.Marshal(r)
+	_ = common.Unmarshal(data, &result)
 	return result
 }
 
@@ -646,4 +646,6 @@ type ResponsesToolsCall struct {
 	Name        string          `json:"name,omitempty"`
 	Description string          `json:"description,omitempty"`
 	Parameters  json.RawMessage `json:"parameters,omitempty"`
+	Function    json.RawMessage `json:"function,omitempty"`
+	Container   json.RawMessage `json:"container,omitempty"`
 }

dto/openai_response.go

@@ -1,6 +1,9 @@
 package dto
 
-import "encoding/json"
+import (
+	"encoding/json"
+	"one-api/types"
+)
 
 type SimpleResponse struct {
 	Usage `json:"usage"`
@@ -28,7 +31,7 @@ type OpenAITextResponse struct {
 	Object  string                     `json:"object"`
 	Created any                        `json:"created"`
 	Choices []OpenAITextResponseChoice `json:"choices"`
-	Error   *OpenAIError               `json:"error,omitempty"`
+	Error   *types.OpenAIError         `json:"error,omitempty"`
 	Usage   `json:"usage"`
 }
 
@@ -201,7 +204,7 @@ type OpenAIResponsesResponse struct {
 	Object             string               `json:"object"`
 	CreatedAt          int                  `json:"created_at"`
 	Status             string               `json:"status"`
-	Error              *OpenAIError         `json:"error,omitempty"`
+	Error              *types.OpenAIError   `json:"error,omitempty"`
 	IncompleteDetails  *IncompleteDetails   `json:"incomplete_details,omitempty"`
 	Instructions       string               `json:"instructions"`
 	MaxOutputTokens    int                  `json:"max_output_tokens"`

dto/pricing.go

@@ -1,26 +1,11 @@
 package dto
 
-type OpenAIModelPermission struct {
-	Id                 string  `json:"id"`
-	Object             string  `json:"object"`
-	Created            int     `json:"created"`
-	AllowCreateEngine  bool    `json:"allow_create_engine"`
-	AllowSampling      bool    `json:"allow_sampling"`
-	AllowLogprobs      bool    `json:"allow_logprobs"`
-	AllowSearchIndices bool    `json:"allow_search_indices"`
-	AllowView          bool    `json:"allow_view"`
-	AllowFineTuning    bool    `json:"allow_fine_tuning"`
-	Organization       string  `json:"organization"`
-	Group              *string `json:"group"`
-	IsBlocking         bool    `json:"is_blocking"`
-}
+import "one-api/constant"
 
 type OpenAIModels struct {
-	Id         string                  `json:"id"`
-	Object     string                  `json:"object"`
-	Created    int                     `json:"created"`
-	OwnedBy    string                  `json:"owned_by"`
-	Permission []OpenAIModelPermission `json:"permission"`
-	Root       string                  `json:"root"`
-	Parent     *string                 `json:"parent"`
+	Id                     string                  `json:"id"`
+	Object                 string                  `json:"object"`
+	Created                int                     `json:"created"`
+	OwnedBy                string                  `json:"owned_by"`
+	SupportedEndpointTypes []constant.EndpointType `json:"supported_endpoint_types"`
 }

dto/realtime.go

@@ -1,5 +1,7 @@
 package dto
 
+import "one-api/types"
+
 const (
 	RealtimeEventTypeError              = "error"
 	RealtimeEventTypeSessionUpdate      = "session.update"
@@ -23,12 +25,12 @@ type RealtimeEvent struct {
 	EventId string `json:"event_id"`
 	Type    string `json:"type"`
 	//PreviousItemId string `json:"previous_item_id"`
-	Session  *RealtimeSession  `json:"session,omitempty"`
-	Item     *RealtimeItem     `json:"item,omitempty"`
-	Error    *OpenAIError      `json:"error,omitempty"`
-	Response *RealtimeResponse `json:"response,omitempty"`
-	Delta    string            `json:"delta,omitempty"`
-	Audio    string            `json:"audio,omitempty"`
+	Session  *RealtimeSession   `json:"session,omitempty"`
+	Item     *RealtimeItem      `json:"item,omitempty"`
+	Error    *types.OpenAIError `json:"error,omitempty"`
+	Response *RealtimeResponse  `json:"response,omitempty"`
+	Delta    string             `json:"delta,omitempty"`
+	Audio    string             `json:"audio,omitempty"`
 }
 
 type RealtimeResponse struct {

dto/rerank.go

@@ -4,7 +4,7 @@ type RerankRequest struct {
 	Documents       []any  `json:"documents"`
 	Query           string `json:"query"`
 	Model           string `json:"model"`
-	TopN            int    `json:"top_n"`
+	TopN            int    `json:"top_n,omitempty"`
 	ReturnDocuments *bool  `json:"return_documents,omitempty"`
 	MaxChunkPerDoc  int    `json:"max_chunk_per_doc,omitempty"`
 	OverLapTokens   int    `json:"overlap_tokens,omitempty"`

dto/user_settings.go

@@ -0,0 +1,16 @@
+package dto
+
+type UserSetting struct {
+	NotifyType            string  `json:"notify_type,omitempty"`                    // QuotaWarningType 额度预警类型
+	QuotaWarningThreshold float64 `json:"quota_warning_threshold,omitempty"`        // QuotaWarningThreshold 额度预警阈值
+	WebhookUrl            string  `json:"webhook_url,omitempty"`                    // WebhookUrl webhook地址
+	WebhookSecret         string  `json:"webhook_secret,omitempty"`                 // WebhookSecret webhook密钥
+	NotificationEmail     string  `json:"notification_email,omitempty"`             // NotificationEmail 通知邮箱地址
+	AcceptUnsetRatioModel bool    `json:"accept_unset_model_ratio_model,omitempty"` // AcceptUnsetRatioModel 是否接受未设置价格的模型
+	RecordIpLog           bool    `json:"record_ip_log,omitempty"`                  // 是否记录请求和错误日志IP
+}
+
+var (
+	NotifyTypeEmail   = "email"   // Email 邮件
+	NotifyTypeWebhook = "webhook" // Webhook
+)

main.go

@@ -32,14 +32,13 @@ var buildFS embed.FS
 var indexPage []byte
 
 func main() {
-	err := godotenv.Load(".env")
+
+	err := InitResources()
 	if err != nil {
-		common.SysLog("Support for .env file is disabled: " + err.Error())
+		common.FatalLog("failed to initialize resources: " + err.Error())
+		return
 	}
 
-	common.LoadEnv()
-
-	common.SetupLogger()
 	common.SysLog("New API " + common.Version + " started")
 	if os.Getenv("GIN_MODE") != "debug" {
 		gin.SetMode(gin.ReleaseMode)
@@ -47,19 +46,7 @@ func main() {
 	if common.DebugEnabled {
 		common.SysLog("running in debug mode")
 	}
-	// Initialize SQL Database
-	err = model.InitDB()
-	if err != nil {
-		common.FatalLog("failed to initialize database: " + err.Error())
-	}
 
-	model.CheckSetup()
-
-	// Initialize SQL Database
-	err = model.InitLogDB()
-	if err != nil {
-		common.FatalLog("failed to initialize database: " + err.Error())
-	}
 	defer func() {
 		err := model.CloseDB()
 		if err != nil {
@@ -67,21 +54,6 @@ func main() {
 		}
 	}()
 
-	// Initialize Redis
-	err = common.InitRedisClient()
-	if err != nil {
-		common.FatalLog("failed to initialize Redis: " + err.Error())
-	}
-
-	// Initialize model settings
-	ratio_setting.InitRatioSettings()
-	// Initialize constants
-	constant.InitEnv()
-	// Initialize options
-	model.InitOptionMap()
-
-	service.InitTokenEncoders()
-
 	if common.RedisEnabled {
 		// for compatibility with old versions
 		common.MemoryCacheEnabled = true
@@ -96,9 +68,9 @@ func main() {
 				if r := recover(); r != nil {
 					common.SysError(fmt.Sprintf("InitChannelCache panic: %v, retrying once", r))
 					// Retry once
-					_, fixErr := model.FixAbility()
+					_, _, fixErr := model.FixAbility()
 					if fixErr != nil {
-						common.SysError(fmt.Sprintf("InitChannelCache failed: %s", fixErr.Error()))
+						common.FatalLog(fmt.Sprintf("InitChannelCache failed: %s", fixErr.Error()))
 					}
 				}
 			}()
@@ -186,3 +158,53 @@ func main() {
 		common.FatalLog("failed to start HTTP server: " + err.Error())
 	}
 }
+
+func InitResources() error {
+	// Initialize resources here if needed
+	// This is a placeholder function for future resource initialization
+	err := godotenv.Load(".env")
+	if err != nil {
+		common.SysLog("未找到 .env 文件，使用默认环境变量，如果需要，请创建 .env 文件并设置相关变量")
+		common.SysLog("No .env file found, using default environment variables. If needed, please create a .env file and set the relevant variables.")
+	}
+
+	common.SetupLogger()
+
+	// 加载环境变量
+	common.InitEnv()
+
+	// Initialize model settings
+	ratio_setting.InitRatioSettings()
+
+	service.InitHttpClient()
+
+	service.InitTokenEncoders()
+
+	// Initialize SQL Database
+	err = model.InitDB()
+	if err != nil {
+		common.FatalLog("failed to initialize database: " + err.Error())
+		return err
+	}
+
+	model.CheckSetup()
+
+	// Initialize options, should after model.InitDB()
+	model.InitOptionMap()
+
+	// 初始化模型
+	model.GetPricing()
+
+	// Initialize SQL Database
+	err = model.InitLogDB()
+	if err != nil {
+		return err
+	}
+
+	// Initialize Redis
+	err = common.InitRedisClient()
+	if err != nil {
+		return err
+	}
+	return nil
+}

middleware/auth.go

@@ -1,6 +1,7 @@
 package middleware
 
 import (
+	"fmt"
 	"net/http"
 	"one-api/common"
 	"one-api/model"
@@ -184,7 +185,7 @@ func TokenAuth() func(c *gin.Context) {
 			}
 		}
 		// gemini api 从query中获取key
-		if strings.HasPrefix(c.Request.URL.Path, "/v1beta/models/") {
+		if strings.HasPrefix(c.Request.URL.Path, "/v1beta/models/") || strings.HasPrefix(c.Request.URL.Path, "/v1/models/") {
 			skKey := c.Query("key")
 			if skKey != "" {
 				c.Request.Header.Set("Authorization", "Bearer "+skKey)
@@ -233,30 +234,41 @@ func TokenAuth() func(c *gin.Context) {
 
 		userCache.WriteContext(c)
 
-		c.Set("id", token.UserId)
-		c.Set("token_id", token.Id)
-		c.Set("token_key", token.Key)
-		c.Set("token_name", token.Name)
-		c.Set("token_unlimited_quota", token.UnlimitedQuota)
-		if !token.UnlimitedQuota {
-			c.Set("token_quota", token.RemainQuota)
-		}
-		if token.ModelLimitsEnabled {
-			c.Set("token_model_limit_enabled", true)
-			c.Set("token_model_limit", token.GetModelLimitsMap())
-		} else {
-			c.Set("token_model_limit_enabled", false)
-		}
-		c.Set("allow_ips", token.GetIpLimitsMap())
-		c.Set("token_group", token.Group)
-		if len(parts) > 1 {
-			if model.IsAdmin(token.UserId) {
-				c.Set("specific_channel_id", parts[1])
-			} else {
-				abortWithOpenAiMessage(c, http.StatusForbidden, "普通用户不支持指定渠道")
-				return
-			}
+		err = SetupContextForToken(c, token, parts...)
+		if err != nil {
+			return
 		}
 		c.Next()
 	}
 }
+
+func SetupContextForToken(c *gin.Context, token *model.Token, parts ...string) error {
+	if token == nil {
+		return fmt.Errorf("token is nil")
+	}
+	c.Set("id", token.UserId)
+	c.Set("token_id", token.Id)
+	c.Set("token_key", token.Key)
+	c.Set("token_name", token.Name)
+	c.Set("token_unlimited_quota", token.UnlimitedQuota)
+	if !token.UnlimitedQuota {
+		c.Set("token_quota", token.RemainQuota)
+	}
+	if token.ModelLimitsEnabled {
+		c.Set("token_model_limit_enabled", true)
+		c.Set("token_model_limit", token.GetModelLimitsMap())
+	} else {
+		c.Set("token_model_limit_enabled", false)
+	}
+	c.Set("allow_ips", token.GetIpLimitsMap())
+	c.Set("token_group", token.Group)
+	if len(parts) > 1 {
+		if model.IsAdmin(token.UserId) {
+			c.Set("specific_channel_id", parts[1])
+		} else {
+			abortWithOpenAiMessage(c, http.StatusForbidden, "普通用户不支持指定渠道")
+			return fmt.Errorf("普通用户不支持指定渠道")
+		}
+	}
+	return nil
+}

middleware/distributor.go

@@ -21,11 +21,12 @@ import (
 
 type ModelRequest struct {
 	Model string `json:"model"`
+	Group string `json:"group,omitempty"`
 }
 
 func Distribute() func(c *gin.Context) {
 	return func(c *gin.Context) {
-		allowIpsMap := c.GetStringMap("allow_ips")
+		allowIpsMap := common.GetContextKeyStringMap(c, constant.ContextKeyTokenAllowIps)
 		if len(allowIpsMap) != 0 {
 			clientIp := c.ClientIP()
 			if _, ok := allowIpsMap[clientIp]; !ok {
@@ -34,14 +35,14 @@ func Distribute() func(c *gin.Context) {
 			}
 		}
 		var channel *model.Channel
-		channelId, ok := c.Get("specific_channel_id")
+		channelId, ok := common.GetContextKey(c, constant.ContextKeyTokenSpecificChannelId)
 		modelRequest, shouldSelectChannel, err := getModelRequest(c)
 		if err != nil {
 			abortWithOpenAiMessage(c, http.StatusBadRequest, "Invalid request, "+err.Error())
 			return
 		}
-		userGroup := c.GetString(constant.ContextKeyUserGroup)
-		tokenGroup := c.GetString("token_group")
+		userGroup := common.GetContextKeyString(c, constant.ContextKeyUserGroup)
+		tokenGroup := common.GetContextKeyString(c, constant.ContextKeyTokenGroup)
 		if tokenGroup != "" {
 			// check common.UserUsableGroups[userGroup]
 			if _, ok := setting.GetUserUsableGroups(userGroup)[tokenGroup]; !ok {
@@ -57,7 +58,7 @@ func Distribute() func(c *gin.Context) {
 			}
 			userGroup = tokenGroup
 		}
-		c.Set(constant.ContextKeyUsingGroup, userGroup)
+		common.SetContextKey(c, constant.ContextKeyUsingGroup, userGroup)
 		if ok {
 			id, err := strconv.Atoi(channelId.(string))
 			if err != nil {
@@ -76,9 +77,9 @@ func Distribute() func(c *gin.Context) {
 		} else {
 			// Select a channel for the user
 			// check token model mapping
-			modelLimitEnable := c.GetBool("token_model_limit_enabled")
+			modelLimitEnable := common.GetContextKeyBool(c, constant.ContextKeyTokenModelLimitEnabled)
 			if modelLimitEnable {
-				s, ok := c.Get("token_model_limit")
+				s, ok := common.GetContextKey(c, constant.ContextKeyTokenModelLimit)
 				var tokenModelLimit map[string]bool
 				if ok {
 					tokenModelLimit = s.(map[string]bool)
@@ -121,7 +122,7 @@ func Distribute() func(c *gin.Context) {
 				}
 			}
 		}
-		c.Set(constant.ContextKeyRequestStartTime, time.Now())
+		common.SetContextKey(c, constant.ContextKeyRequestStartTime, time.Now())
 		SetupContextForSelectedChannel(c, channel, modelRequest.Model)
 		c.Next()
 	}
@@ -171,15 +172,25 @@ func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
 		c.Set("platform", string(constant.TaskPlatformSuno))
 		c.Set("relay_mode", relayMode)
 	} else if strings.Contains(c.Request.URL.Path, "/v1/video/generations") {
-		relayMode := relayconstant.Path2RelayKling(c.Request.Method, c.Request.URL.Path)
-		if relayMode == relayconstant.RelayModeKlingFetchByID {
-			shouldSelectChannel = false
+		err = common.UnmarshalBodyReusable(c, &modelRequest)
+		var platform string
+		var relayMode int
+		if strings.HasPrefix(modelRequest.Model, "jimeng") {
+			platform = string(constant.TaskPlatformJimeng)
+			relayMode = relayconstant.Path2RelayJimeng(c.Request.Method, c.Request.URL.Path)
+			if relayMode == relayconstant.RelayModeJimengFetchByID {
+				shouldSelectChannel = false
+			}
 		} else {
-			err = common.UnmarshalBodyReusable(c, &modelRequest)
+			platform = string(constant.TaskPlatformKling)
+			relayMode = relayconstant.Path2RelayKling(c.Request.Method, c.Request.URL.Path)
+			if relayMode == relayconstant.RelayModeKlingFetchByID {
+				shouldSelectChannel = false
+			}
 		}
-		c.Set("platform", string(constant.TaskPlatformKling))
+		c.Set("platform", platform)
 		c.Set("relay_mode", relayMode)
-	} else if strings.HasPrefix(c.Request.URL.Path, "/v1beta/models/") {
+	} else if strings.HasPrefix(c.Request.URL.Path, "/v1beta/models/") || strings.HasPrefix(c.Request.URL.Path, "/v1/models/") {
 		// Gemini API 路径处理: /v1beta/models/gemini-2.0-flash:generateContent
 		relayMode := relayconstant.RelayModeGemini
 		modelName := extractModelNameFromGeminiPath(c.Request.URL.Path)
@@ -227,6 +238,14 @@ func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
 		}
 		c.Set("relay_mode", relayMode)
 	}
+	if strings.HasPrefix(c.Request.URL.Path, "/pg/chat/completions") {
+		// playground chat completions
+		err = common.UnmarshalBodyReusable(c, &modelRequest)
+		if err != nil {
+			return nil, false, errors.New("无效的请求, " + err.Error())
+		}
+		common.SetContextKey(c, constant.ContextKeyTokenGroup, modelRequest.Group)
+	}
 	return &modelRequest, shouldSelectChannel, nil
 }
 
@@ -235,37 +254,42 @@ func SetupContextForSelectedChannel(c *gin.Context, channel *model.Channel, mode
 	if channel == nil {
 		return
 	}
-	c.Set("channel_id", channel.Id)
-	c.Set("channel_name", channel.Name)
-	c.Set("channel_type", channel.Type)
-	c.Set("channel_create_time", channel.CreatedTime)
-	c.Set("channel_setting", channel.GetSetting())
-	c.Set("param_override", channel.GetParamOverride())
-	if nil != channel.OpenAIOrganization && "" != *channel.OpenAIOrganization {
-		c.Set("channel_organization", *channel.OpenAIOrganization)
+	common.SetContextKey(c, constant.ContextKeyChannelId, channel.Id)
+	common.SetContextKey(c, constant.ContextKeyChannelName, channel.Name)
+	common.SetContextKey(c, constant.ContextKeyChannelType, channel.Type)
+	common.SetContextKey(c, constant.ContextKeyChannelCreateTime, channel.CreatedTime)
+	common.SetContextKey(c, constant.ContextKeyChannelSetting, channel.GetSetting())
+	common.SetContextKey(c, constant.ContextKeyChannelParamOverride, channel.GetParamOverride())
+	if nil != channel.OpenAIOrganization && *channel.OpenAIOrganization != "" {
+		common.SetContextKey(c, constant.ContextKeyChannelOrganization, *channel.OpenAIOrganization)
+	}
+	common.SetContextKey(c, constant.ContextKeyChannelAutoBan, channel.GetAutoBan())
+	common.SetContextKey(c, constant.ContextKeyChannelModelMapping, channel.GetModelMapping())
+	common.SetContextKey(c, constant.ContextKeyChannelStatusCodeMapping, channel.GetStatusCodeMapping())
+	if channel.ChannelInfo.IsMultiKey {
+		common.SetContextKey(c, constant.ContextKeyChannelIsMultiKey, true)
+
 	}
-	c.Set("auto_ban", channel.GetAutoBan())
-	c.Set("model_mapping", channel.GetModelMapping())
-	c.Set("status_code_mapping", channel.GetStatusCodeMapping())
 	c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
-	c.Set("base_url", channel.GetBaseURL())
+	common.SetContextKey(c, constant.ContextKeyChannelBaseUrl, channel.GetBaseURL())
+
 	// TODO: api_version统一
 	switch channel.Type {
-	case common.ChannelTypeAzure:
+	case constant.ChannelTypeAzure:
 		c.Set("api_version", channel.Other)
-	case common.ChannelTypeVertexAi:
+	case constant.ChannelTypeVertexAi:
 		c.Set("region", channel.Other)
-	case common.ChannelTypeXunfei:
+	case constant.ChannelTypeXunfei:
 		c.Set("api_version", channel.Other)
-	case common.ChannelTypeGemini:
+	case constant.ChannelTypeGemini:
 		c.Set("api_version", channel.Other)
-	case common.ChannelTypeAli:
+	case constant.ChannelTypeAli:
 		c.Set("plugin", channel.Other)
-	case common.ChannelCloudflare:
+	case constant.ChannelCloudflare:
 		c.Set("api_version", channel.Other)
-	case common.ChannelTypeMokaAI:
+	case constant.ChannelTypeMokaAI:
 		c.Set("api_version", channel.Other)
-	case common.ChannelTypeCoze:
+	case constant.ChannelTypeCoze:
 		c.Set("bot_id", channel.Other)
 	}
 }

middleware/kling_adapter.go

@@ -0,0 +1,47 @@
+package middleware
+
+import (
+	"bytes"
+	"encoding/json"
+	"io"
+	"one-api/common"
+	"one-api/constant"
+
+	"github.com/gin-gonic/gin"
+)
+
+func KlingRequestConvert() func(c *gin.Context) {
+	return func(c *gin.Context) {
+		var originalReq map[string]interface{}
+		if err := common.UnmarshalBodyReusable(c, &originalReq); err != nil {
+			c.Next()
+			return
+		}
+
+		model, _ := originalReq["model"].(string)
+		prompt, _ := originalReq["prompt"].(string)
+
+		unifiedReq := map[string]interface{}{
+			"model":    model,
+			"prompt":   prompt,
+			"metadata": originalReq,
+		}
+
+		jsonData, err := json.Marshal(unifiedReq)
+		if err != nil {
+			c.Next()
+			return
+		}
+
+		// Rewrite request body and path
+		c.Request.Body = io.NopCloser(bytes.NewBuffer(jsonData))
+		c.Request.URL.Path = "/v1/video/generations"
+		if image := originalReq["image"]; image == "" {
+			c.Set("action", constant.TaskActionTextGenerate)
+		}
+
+		// We have to reset the request body for the next handlers
+		c.Set(common.KeyRequestBody, jsonData)
+		c.Next()
+	}
+}

middleware/model-rate-limit.go

@@ -177,9 +177,9 @@ func ModelRequestRateLimit() func(c *gin.Context) {
 		successMaxCount := setting.ModelRequestRateLimitSuccessCount
 
 		// 获取分组
-		group := c.GetString("token_group")
+		group := common.GetContextKeyString(c, constant.ContextKeyTokenGroup)
 		if group == "" {
-			group = c.GetString(constant.ContextKeyUserGroup)
+			group = common.GetContextKeyString(c, constant.ContextKeyUserGroup)
 		}
 
 		//获取分组的限流配置

model/ability.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"one-api/common"
 	"strings"
+	"sync"
 
 	"github.com/samber/lo"
 	"gorm.io/gorm"
@@ -21,7 +22,22 @@ type Ability struct {
 	Tag       *string `json:"tag" gorm:"index"`
 }
 
-func GetGroupModels(group string) []string {
+type AbilityWithChannel struct {
+	Ability
+	ChannelType int `json:"channel_type"`
+}
+
+func GetAllEnableAbilityWithChannels() ([]AbilityWithChannel, error) {
+	var abilities []AbilityWithChannel
+	err := DB.Table("abilities").
+		Select("abilities.*, channels.type as channel_type").
+		Joins("left join channels on abilities.channel_id = channels.id").
+		Where("abilities.enabled = ?", true).
+		Scan(&abilities).Error
+	return abilities, err
+}
+
+func GetGroupEnabledModels(group string) []string {
 	var models []string
 	// Find distinct models
 	DB.Table("abilities").Where(commonGroupCol+" = ? and enabled = ?", group, true).Distinct("model").Pluck("model", &models)
@@ -46,7 +62,7 @@ func getPriority(group string, model string, retry int) (int, error) {
 	var priorities []int
 	err := DB.Model(&Ability{}).
 		Select("DISTINCT(priority)").
-		Where(commonGroupCol+" = ? and model = ? and enabled = ?", group, model, commonTrueVal).
+		Where(commonGroupCol+" = ? and model = ? and enabled = ?", group, model, true).
 		Order("priority DESC").              // 按优先级降序排序
 		Pluck("priority", &priorities).Error // Pluck用于将查询的结果直接扫描到一个切片中
 
@@ -72,14 +88,14 @@ func getPriority(group string, model string, retry int) (int, error) {
 }
 
 func getChannelQuery(group string, model string, retry int) *gorm.DB {
-	maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(commonGroupCol+" = ? and model = ? and enabled = ?", group, model, commonTrueVal)
-	channelQuery := DB.Where(commonGroupCol+" = ? and model = ? and enabled = ? and priority = (?)", group, model, commonTrueVal, maxPrioritySubQuery)
+	maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(commonGroupCol+" = ? and model = ? and enabled = ?", group, model, true)
+	channelQuery := DB.Where(commonGroupCol+" = ? and model = ? and enabled = ? and priority = (?)", group, model, true, maxPrioritySubQuery)
 	if retry != 0 {
 		priority, err := getPriority(group, model, retry)
 		if err != nil {
 			common.SysError(fmt.Sprintf("Get priority failed: %s", err.Error()))
 		} else {
-			channelQuery = DB.Where(commonGroupCol+" = ? and model = ? and enabled = ? and priority = ?", group, model, commonTrueVal, priority)
+			channelQuery = DB.Where(commonGroupCol+" = ? and model = ? and enabled = ? and priority = ?", group, model, true, priority)
 		}
 	}
 
@@ -257,74 +273,45 @@ func UpdateAbilityByTag(tag string, newTag *string, priority *int64, weight *uin
 	return DB.Model(&Ability{}).Where("tag = ?", tag).Updates(ability).Error
 }
 
-func FixAbility() (int, error) {
-	var channelIds []int
-	count := 0
-	// Find all channel ids from channel table
-	err := DB.Model(&Channel{}).Pluck("id", &channelIds).Error
+var fixLock = sync.Mutex{}
+
+func FixAbility() (int, int, error) {
+	lock := fixLock.TryLock()
+	if !lock {
+		return 0, 0, errors.New("已经有一个修复任务在运行中，请稍后再试")
+	}
+	defer fixLock.Unlock()
+	var channels []*Channel
+	// Find all channels
+	err := DB.Model(&Channel{}).Find(&channels).Error
 	if err != nil {
-		common.SysError(fmt.Sprintf("Get channel ids from channel table failed: %s", err.Error()))
-		return 0, err
+		return 0, 0, err
 	}
-
-	// Delete abilities of channels that are not in channel table - in batches to avoid too many placeholders
-	if len(channelIds) > 0 {
-		// Process deletion in chunks to avoid "too many placeholders" error
-		for _, chunk := range lo.Chunk(channelIds, 100) {
-			err = DB.Where("channel_id NOT IN (?)", chunk).Delete(&Ability{}).Error
-			if err != nil {
-				common.SysError(fmt.Sprintf("Delete abilities of channels (batch) that are not in channel table failed: %s", err.Error()))
-				return 0, err
-			}
-		}
-	} else {
-		// If no channels exist, delete all abilities
-		err = DB.Delete(&Ability{}).Error
+	if len(channels) == 0 {
+		return 0, 0, nil
+	}
+	successCount := 0
+	failCount := 0
+	for _, chunk := range lo.Chunk(channels, 50) {
+		ids := lo.Map(chunk, func(c *Channel, _ int) int { return c.Id })
+		// Delete all abilities of this channel
+		err = DB.Where("channel_id IN ?", ids).Delete(&Ability{}).Error
 		if err != nil {
-			common.SysError(fmt.Sprintf("Delete all abilities failed: %s", err.Error()))
-			return 0, err
+			common.SysError(fmt.Sprintf("Delete abilities failed: %s", err.Error()))
+			failCount += len(chunk)
+			continue
 		}
-		common.SysLog("Delete all abilities successfully")
-		return 0, nil
-	}
-
-	common.SysLog(fmt.Sprintf("Delete abilities of channels that are not in channel table successfully, ids: %v", channelIds))
-	count += len(channelIds)
-
-	// Use channelIds to find channel not in abilities table
-	var abilityChannelIds []int
-	err = DB.Table("abilities").Distinct("channel_id").Pluck("channel_id", &abilityChannelIds).Error
-	if err != nil {
-		common.SysError(fmt.Sprintf("Get channel ids from abilities table failed: %s", err.Error()))
-		return count, err
-	}
-
-	var channels []Channel
-	if len(abilityChannelIds) == 0 {
-		err = DB.Find(&channels).Error
-	} else {
-		// Process query in chunks to avoid "too many placeholders" error
-		err = nil
-		for _, chunk := range lo.Chunk(abilityChannelIds, 100) {
-			var channelsChunk []Channel
-			err = DB.Where("id NOT IN (?)", chunk).Find(&channelsChunk).Error
+		// Then add new abilities
+		for _, channel := range chunk {
+			err = channel.AddAbilities()
 			if err != nil {
-				common.SysError(fmt.Sprintf("Find channels not in abilities table failed: %s", err.Error()))
-				return count, err
+				common.SysError(fmt.Sprintf("Add abilities for channel %d failed: %s", channel.Id, err.Error()))
+				failCount++
+			} else {
+				successCount++
 			}
-			channels = append(channels, channelsChunk...)
-		}
-	}
-
-	for _, channel := range channels {
-		err := channel.UpdateAbilities(nil)
-		if err != nil {
-			common.SysError(fmt.Sprintf("Update abilities of channel %d failed: %s", channel.Id, err.Error()))
-		} else {
-			common.SysLog(fmt.Sprintf("Update abilities of channel %d successfully", channel.Id))
-			count++
 		}
 	}
 	InitChannelCache()
-	return count, nil
+	return successCount, failCount, nil
 }

model/channel.go

@@ -1,8 +1,13 @@
 package model
 
 import (
+	"database/sql/driver"
 	"encoding/json"
+	"fmt"
+	"math/rand"
 	"one-api/common"
+	"one-api/constant"
+	"one-api/dto"
 	"strings"
 	"sync"
 
@@ -35,8 +40,100 @@ type Channel struct {
 	AutoBan           *int    `json:"auto_ban" gorm:"default:1"`
 	OtherInfo         string  `json:"other_info"`
 	Tag               *string `json:"tag" gorm:"index"`
-	Setting           *string `json:"setting" gorm:"type:text"`
+	Setting           *string `json:"setting" gorm:"type:text"` // 渠道额外设置
 	ParamOverride     *string `json:"param_override" gorm:"type:text"`
+	// add after v0.8.5
+	ChannelInfo ChannelInfo `json:"channel_info" gorm:"type:json"`
+}
+
+type ChannelInfo struct {
+	IsMultiKey           bool                  `json:"is_multi_key"`            // 是否多Key模式
+	MultiKeyStatusList   map[int]int           `json:"multi_key_status_list"`   // key状态列表，key index -> status
+	MultiKeyPollingIndex int                   `json:"multi_key_polling_index"` // 多Key模式下轮询的key索引
+	MultiKeyMode         constant.MultiKeyMode `json:"multi_key_mode"`
+}
+
+// Value implements driver.Valuer interface
+func (c ChannelInfo) Value() (driver.Value, error) {
+	return common.Marshal(&c)
+}
+
+// Scan implements sql.Scanner interface
+func (c *ChannelInfo) Scan(value interface{}) error {
+	bytesValue, _ := value.([]byte)
+	return common.Unmarshal(bytesValue, c)
+}
+
+func (channel *Channel) getKeys() []string {
+	if channel.Key == "" {
+		return []string{}
+	}
+	// use \n to split keys
+	keys := strings.Split(strings.Trim(channel.Key, "\n"), "\n")
+	return keys
+}
+
+func (channel *Channel) GetNextEnabledKey() (string, error) {
+	// If not in multi-key mode, return the original key string directly.
+	if !channel.ChannelInfo.IsMultiKey {
+		return channel.Key, nil
+	}
+
+	// Obtain all keys (split by \n)
+	keys := channel.getKeys()
+	if len(keys) == 0 {
+		// No keys available, return error, should disable the channel
+		return "", fmt.Errorf("no valid keys in channel")
+	}
+
+	statusList := channel.ChannelInfo.MultiKeyStatusList
+	// helper to get key status, default to enabled when missing
+	getStatus := func(idx int) int {
+		if statusList == nil {
+			return common.ChannelStatusEnabled
+		}
+		if status, ok := statusList[idx]; ok {
+			return status
+		}
+		return common.ChannelStatusEnabled
+	}
+
+	// Collect indexes of enabled keys
+	enabledIdx := make([]int, 0, len(keys))
+	for i := range keys {
+		if getStatus(i) == common.ChannelStatusEnabled {
+			enabledIdx = append(enabledIdx, i)
+		}
+	}
+	// If no specific status list or none enabled, fall back to first key
+	if len(enabledIdx) == 0 {
+		return keys[0], nil
+	}
+
+	switch channel.ChannelInfo.MultiKeyMode {
+	case constant.MultiKeyModeRandom:
+		// Randomly pick one enabled key
+		return keys[enabledIdx[rand.Intn(len(enabledIdx))]], nil
+	case constant.MultiKeyModePolling:
+		// Start from the saved polling index and look for the next enabled key
+		start := channel.ChannelInfo.MultiKeyPollingIndex
+		if start < 0 || start >= len(keys) {
+			start = 0
+		}
+		for i := 0; i < len(keys); i++ {
+			idx := (start + i) % len(keys)
+			if getStatus(idx) == common.ChannelStatusEnabled {
+				// update polling index for next call (point to the next position)
+				channel.ChannelInfo.MultiKeyPollingIndex = (idx + 1) % len(keys)
+				return keys[idx], nil
+			}
+		}
+		// Fallback – should not happen, but return first enabled key
+		return keys[enabledIdx[0]], nil
+	default:
+		// Unknown mode, default to first enabled key (or original key string)
+		return keys[enabledIdx[0]], nil
+	}
 }
 
 func (channel *Channel) GetModels() []string {
@@ -514,8 +611,19 @@ func SearchTags(keyword string, group string, model string, idSort bool) ([]*str
 	return tags, nil
 }
 
-func (channel *Channel) GetSetting() map[string]interface{} {
-	setting := make(map[string]interface{})
+func (channel *Channel) ValidateSettings() error {
+	channelParams := &dto.ChannelSettings{}
+	if channel.Setting != nil && *channel.Setting != "" {
+		err := json.Unmarshal([]byte(*channel.Setting), channelParams)
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func (channel *Channel) GetSetting() dto.ChannelSettings {
+	setting := dto.ChannelSettings{}
 	if channel.Setting != nil && *channel.Setting != "" {
 		err := json.Unmarshal([]byte(*channel.Setting), &setting)
 		if err != nil {
@@ -525,7 +633,7 @@ func (channel *Channel) GetSetting() map[string]interface{} {
 	return setting
 }
 
-func (channel *Channel) SetSetting(setting map[string]interface{}) {
+func (channel *Channel) SetSetting(setting dto.ChannelSettings) {
 	settingBytes, err := json.Marshal(setting)
 	if err != nil {
 		common.SysError("failed to marshal setting: " + err.Error())

model/log.go

@@ -4,7 +4,6 @@ import (
 	"context"
 	"fmt"
 	"one-api/common"
-	"one-api/constant"
 	"os"
 	"strings"
 	"time"
@@ -50,7 +49,7 @@ func formatUserLogs(logs []*Log) {
 	for i := range logs {
 		logs[i].ChannelName = ""
 		var otherMap map[string]interface{}
-		otherMap = common.StrToMap(logs[i].Other)
+		otherMap, _ = common.StrToMap(logs[i].Other)
 		if otherMap != nil {
 			// delete admin
 			delete(otherMap, "admin_info")
@@ -100,10 +99,8 @@ func RecordErrorLog(c *gin.Context, userId int, channelId int, modelName string,
 	// 判断是否需要记录 IP
 	needRecordIp := false
 	if settingMap, err := GetUserSetting(userId, false); err == nil {
-		if v, ok := settingMap[constant.UserSettingRecordIpLog]; ok {
-			if vb, ok := v.(bool); ok && vb {
-				needRecordIp = true
-			}
+		if settingMap.RecordIpLog {
+			needRecordIp = true
 		}
 	}
 	log := &Log{
@@ -136,22 +133,34 @@ func RecordErrorLog(c *gin.Context, userId int, channelId int, modelName string,
 	}
 }
 
-func RecordConsumeLog(c *gin.Context, userId int, channelId int, promptTokens int, completionTokens int,
-	modelName string, tokenName string, quota int, content string, tokenId int, userQuota int, useTimeSeconds int,
-	isStream bool, group string, other map[string]interface{}) {
-	common.LogInfo(c, fmt.Sprintf("record consume log: userId=%d, 用户调用前余额=%d, channelId=%d, promptTokens=%d, completionTokens=%d, modelName=%s, tokenName=%s, quota=%d, content=%s", userId, userQuota, channelId, promptTokens, completionTokens, modelName, tokenName, quota, content))
+type RecordConsumeLogParams struct {
+	ChannelId        int                    `json:"channel_id"`
+	PromptTokens     int                    `json:"prompt_tokens"`
+	CompletionTokens int                    `json:"completion_tokens"`
+	ModelName        string                 `json:"model_name"`
+	TokenName        string                 `json:"token_name"`
+	Quota            int                    `json:"quota"`
+	Content          string                 `json:"content"`
+	TokenId          int                    `json:"token_id"`
+	UserQuota        int                    `json:"user_quota"`
+	UseTimeSeconds   int                    `json:"use_time_seconds"`
+	IsStream         bool                   `json:"is_stream"`
+	Group            string                 `json:"group"`
+	Other            map[string]interface{} `json:"other"`
+}
+
+func RecordConsumeLog(c *gin.Context, userId int, params RecordConsumeLogParams) {
+	common.LogInfo(c, fmt.Sprintf("record consume log: userId=%d, params=%s", userId, common.GetJsonString(params)))
 	if !common.LogConsumeEnabled {
 		return
 	}
 	username := c.GetString("username")
-	otherStr := common.MapToJsonStr(other)
+	otherStr := common.MapToJsonStr(params.Other)
 	// 判断是否需要记录 IP
 	needRecordIp := false
 	if settingMap, err := GetUserSetting(userId, false); err == nil {
-		if v, ok := settingMap[constant.UserSettingRecordIpLog]; ok {
-			if vb, ok := v.(bool); ok && vb {
-				needRecordIp = true
-			}
+		if settingMap.RecordIpLog {
+			needRecordIp = true
 		}
 	}
 	log := &Log{
@@ -159,17 +168,17 @@ func RecordConsumeLog(c *gin.Context, userId int, channelId int, promptTokens in
 		Username:         username,
 		CreatedAt:        common.GetTimestamp(),
 		Type:             LogTypeConsume,
-		Content:          content,
-		PromptTokens:     promptTokens,
-		CompletionTokens: completionTokens,
-		TokenName:        tokenName,
-		ModelName:        modelName,
-		Quota:            quota,
-		ChannelId:        channelId,
-		TokenId:          tokenId,
-		UseTime:          useTimeSeconds,
-		IsStream:         isStream,
-		Group:            group,
+		Content:          params.Content,
+		PromptTokens:     params.PromptTokens,
+		CompletionTokens: params.CompletionTokens,
+		TokenName:        params.TokenName,
+		ModelName:        params.ModelName,
+		Quota:            params.Quota,
+		ChannelId:        params.ChannelId,
+		TokenId:          params.TokenId,
+		UseTime:          params.UseTimeSeconds,
+		IsStream:         params.IsStream,
+		Group:            params.Group,
 		Ip: func() string {
 			if needRecordIp {
 				return c.ClientIP()
@@ -184,7 +193,7 @@ func RecordConsumeLog(c *gin.Context, userId int, channelId int, promptTokens in
 	}
 	if common.DataExportEnabled {
 		gopool.Go(func() {
-			LogQuotaData(userId, username, modelName, quota, common.GetTimestamp(), promptTokens+completionTokens)
+			LogQuotaData(userId, username, params.ModelName, params.Quota, common.GetTimestamp(), params.PromptTokens+params.CompletionTokens)
 		})
 	}
 }

model/main.go

@@ -57,7 +57,7 @@ func initCol() {
 		}
 	}
 	// log sql type and database type
-	common.SysLog("Using Log SQL Type: " + common.LogSqlType)
+	//common.SysLog("Using Log SQL Type: " + common.LogSqlType)
 }
 
 var DB *gorm.DB
@@ -225,12 +225,6 @@ func InitLogDB() (err error) {
 		if !common.IsMasterNode {
 			return nil
 		}
-		//if common.UsingMySQL {
-		//	_, _ = sqlDB.Exec("DROP INDEX idx_channels_key ON channels;")             // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY action VARCHAR(40);")   // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY progress VARCHAR(30);") // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY status VARCHAR(20);")   // TODO: delete this line when most users have upgraded
-		//}
 		common.SysLog("database migration started")
 		err = migrateLOGDB()
 		return err

model/midjourney.go

@@ -14,6 +14,8 @@ type Midjourney struct {
 	StartTime   int64  `json:"start_time" gorm:"index"`
 	FinishTime  int64  `json:"finish_time" gorm:"index"`
 	ImageUrl    string `json:"image_url"`
+	VideoUrl    string `json:"video_url"`
+	VideoUrls   string `json:"video_urls"`
 	Status      string `json:"status" gorm:"type:varchar(20);index"`
 	Progress    string `json:"progress" gorm:"type:varchar(30);index"`
 	FailReason  string `json:"fail_reason"`

model/pricing.go

@@ -1,20 +1,24 @@
 package model
 
 import (
+	"fmt"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/setting/ratio_setting"
+	"one-api/types"
 	"sync"
 	"time"
 )
 
 type Pricing struct {
-	ModelName       string   `json:"model_name"`
-	QuotaType       int      `json:"quota_type"`
-	ModelRatio      float64  `json:"model_ratio"`
-	ModelPrice      float64  `json:"model_price"`
-	OwnerBy         string   `json:"owner_by"`
-	CompletionRatio float64  `json:"completion_ratio"`
-	EnableGroup     []string `json:"enable_groups,omitempty"`
+	ModelName              string                  `json:"model_name"`
+	QuotaType              int                     `json:"quota_type"`
+	ModelRatio             float64                 `json:"model_ratio"`
+	ModelPrice             float64                 `json:"model_price"`
+	OwnerBy                string                  `json:"owner_by"`
+	CompletionRatio        float64                 `json:"completion_ratio"`
+	EnableGroup            []string                `json:"enable_groups"`
+	SupportedEndpointTypes []constant.EndpointType `json:"supported_endpoint_types"`
 }
 
 var (
@@ -23,47 +27,89 @@ var (
 	updatePricingLock  sync.Mutex
 )
 
-func GetPricing() []Pricing {
-	updatePricingLock.Lock()
-	defer updatePricingLock.Unlock()
+var (
+	modelSupportEndpointTypes = make(map[string][]constant.EndpointType)
+	modelSupportEndpointsLock = sync.RWMutex{}
+)
 
+func GetPricing() []Pricing {
 	if time.Since(lastGetPricingTime) > time.Minute*1 || len(pricingMap) == 0 {
-		updatePricing()
+		updatePricingLock.Lock()
+		defer updatePricingLock.Unlock()
+		// Double check after acquiring the lock
+		if time.Since(lastGetPricingTime) > time.Minute*1 || len(pricingMap) == 0 {
+			modelSupportEndpointsLock.Lock()
+			defer modelSupportEndpointsLock.Unlock()
+			updatePricing()
+		}
 	}
-	//if group != "" {
-	//	userPricingMap := make([]Pricing, 0)
-	//	models := GetGroupModels(group)
-	//	for _, pricing := range pricingMap {
-	//		if !common.StringsContains(models, pricing.ModelName) {
-	//			pricing.Available = false
-	//		}
-	//		userPricingMap = append(userPricingMap, pricing)
-	//	}
-	//	return userPricingMap
-	//}
 	return pricingMap
 }
 
+func GetModelSupportEndpointTypes(model string) []constant.EndpointType {
+	if model == "" {
+		return make([]constant.EndpointType, 0)
+	}
+	modelSupportEndpointsLock.RLock()
+	defer modelSupportEndpointsLock.RUnlock()
+	if endpoints, ok := modelSupportEndpointTypes[model]; ok {
+		return endpoints
+	}
+	return make([]constant.EndpointType, 0)
+}
+
 func updatePricing() {
 	//modelRatios := common.GetModelRatios()
-	enableAbilities := GetAllEnableAbilities()
-	modelGroupsMap := make(map[string][]string)
+	enableAbilities, err := GetAllEnableAbilityWithChannels()
+	if err != nil {
+		common.SysError(fmt.Sprintf("GetAllEnableAbilityWithChannels error: %v", err))
+		return
+	}
+	modelGroupsMap := make(map[string]*types.Set[string])
+
 	for _, ability := range enableAbilities {
-		groups := modelGroupsMap[ability.Model]
-		if groups == nil {
-			groups = make([]string, 0)
+		groups, ok := modelGroupsMap[ability.Model]
+		if !ok {
+			groups = types.NewSet[string]()
+			modelGroupsMap[ability.Model] = groups
 		}
-		if !common.StringsContains(groups, ability.Group) {
-			groups = append(groups, ability.Group)
+		groups.Add(ability.Group)
+	}
+
+	//这里使用切片而不是Set，因为一个模型可能支持多个端点类型，并且第一个端点是优先使用端点
+	modelSupportEndpointsStr := make(map[string][]string)
+
+	for _, ability := range enableAbilities {
+		endpoints, ok := modelSupportEndpointsStr[ability.Model]
+		if !ok {
+			endpoints = make([]string, 0)
+			modelSupportEndpointsStr[ability.Model] = endpoints
 		}
-		modelGroupsMap[ability.Model] = groups
+		channelTypes := common.GetEndpointTypesByChannelType(ability.ChannelType, ability.Model)
+		for _, channelType := range channelTypes {
+			if !common.StringsContains(endpoints, string(channelType)) {
+				endpoints = append(endpoints, string(channelType))
+			}
+		}
+		modelSupportEndpointsStr[ability.Model] = endpoints
+	}
+
+	modelSupportEndpointTypes = make(map[string][]constant.EndpointType)
+	for model, endpoints := range modelSupportEndpointsStr {
+		supportedEndpoints := make([]constant.EndpointType, 0)
+		for _, endpointStr := range endpoints {
+			endpointType := constant.EndpointType(endpointStr)
+			supportedEndpoints = append(supportedEndpoints, endpointType)
+		}
+		modelSupportEndpointTypes[model] = supportedEndpoints
 	}
 
 	pricingMap = make([]Pricing, 0)
 	for model, groups := range modelGroupsMap {
 		pricing := Pricing{
-			ModelName:   model,
-			EnableGroup: groups,
+			ModelName:              model,
+			EnableGroup:            groups.Items(),
+			SupportedEndpointTypes: modelSupportEndpointTypes[model],
 		}
 		modelPrice, findPrice := ratio_setting.GetModelPrice(model, false)
 		if findPrice {

model/token_cache.go

@@ -10,7 +10,7 @@ import (
 func cacheSetToken(token Token) error {
 	key := common.GenerateHMAC(token.Key)
 	token.Clean()
-	err := common.RedisHSetObj(fmt.Sprintf("token:%s", key), &token, time.Duration(constant.RedisKeyCacheSeconds())*time.Second)
+	err := common.RedisHSetObj(fmt.Sprintf("token:%s", key), &token, time.Duration(common.RedisKeyCacheSeconds())*time.Second)
 	if err != nil {
 		return err
 	}

model/user.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"fmt"
 	"one-api/common"
+	"one-api/dto"
 	"strconv"
 	"strings"
 
@@ -68,14 +69,18 @@ func (user *User) SetAccessToken(token string) {
 	user.AccessToken = &token
 }
 
-func (user *User) GetSetting() map[string]interface{} {
-	if user.Setting == "" {
-		return nil
+func (user *User) GetSetting() dto.UserSetting {
+	setting := dto.UserSetting{}
+	if user.Setting != "" {
+		err := json.Unmarshal([]byte(user.Setting), &setting)
+		if err != nil {
+			common.SysError("failed to unmarshal setting: " + err.Error())
+		}
 	}
-	return common.StrToMap(user.Setting)
+	return setting
 }
 
-func (user *User) SetSetting(setting map[string]interface{}) {
+func (user *User) SetSetting(setting dto.UserSetting) {
 	settingBytes, err := json.Marshal(setting)
 	if err != nil {
 		common.SysError("failed to marshal setting: " + err.Error())
@@ -114,7 +119,7 @@ func GetMaxUserId() int {
 	return user.Id
 }
 
-func GetAllUsers(startIdx int, num int) (users []*User, total int64, err error) {
+func GetAllUsers(pageInfo *common.PageInfo) (users []*User, total int64, err error) {
 	// Start transaction
 	tx := DB.Begin()
 	if tx.Error != nil {
@@ -134,7 +139,7 @@ func GetAllUsers(startIdx int, num int) (users []*User, total int64, err error)
 	}
 
 	// Get paginated users within same transaction
-	err = tx.Unscoped().Order("id desc").Limit(num).Offset(startIdx).Omit("password").Find(&users).Error
+	err = tx.Unscoped().Order("id desc").Limit(pageInfo.GetPageSize()).Offset(pageInfo.GetStartIdx()).Omit("password").Find(&users).Error
 	if err != nil {
 		tx.Rollback()
 		return nil, 0, err
@@ -626,7 +631,7 @@ func GetUserGroup(id int, fromDB bool) (group string, err error) {
 }
 
 // GetUserSetting gets setting from Redis first, falls back to DB if needed
-func GetUserSetting(id int, fromDB bool) (settingMap map[string]interface{}, err error) {
+func GetUserSetting(id int, fromDB bool) (settingMap dto.UserSetting, err error) {
 	var setting string
 	defer func() {
 		// Update Redis cache asynchronously on successful DB read
@@ -648,10 +653,12 @@ func GetUserSetting(id int, fromDB bool) (settingMap map[string]interface{}, err
 	fromDB = true
 	err = DB.Model(&User{}).Where("id = ?", id).Select("setting").Find(&setting).Error
 	if err != nil {
-		return map[string]interface{}{}, err
+		return settingMap, err
 	}
-
-	return common.StrToMap(setting), nil
+	userBase := &UserBase{
+		Setting: setting,
+	}
+	return userBase.GetSetting(), nil
 }
 
 func IncreaseUserQuota(id int, quota int, db bool) (err error) {

model/user_cache.go

@@ -1,10 +1,10 @@
 package model
 
 import (
-	"encoding/json"
 	"fmt"
 	"one-api/common"
 	"one-api/constant"
+	"one-api/dto"
 	"time"
 
 	"github.com/gin-gonic/gin"
@@ -24,28 +24,23 @@ type UserBase struct {
 }
 
 func (user *UserBase) WriteContext(c *gin.Context) {
-	c.Set(constant.ContextKeyUserGroup, user.Group)
-	c.Set(constant.ContextKeyUserQuota, user.Quota)
-	c.Set(constant.ContextKeyUserStatus, user.Status)
-	c.Set(constant.ContextKeyUserEmail, user.Email)
-	c.Set("username", user.Username)
-	c.Set(constant.ContextKeyUserSetting, user.GetSetting())
+	common.SetContextKey(c, constant.ContextKeyUserGroup, user.Group)
+	common.SetContextKey(c, constant.ContextKeyUserQuota, user.Quota)
+	common.SetContextKey(c, constant.ContextKeyUserStatus, user.Status)
+	common.SetContextKey(c, constant.ContextKeyUserEmail, user.Email)
+	common.SetContextKey(c, constant.ContextKeyUserName, user.Username)
+	common.SetContextKey(c, constant.ContextKeyUserSetting, user.GetSetting())
 }
 
-func (user *UserBase) GetSetting() map[string]interface{} {
-	if user.Setting == "" {
-		return nil
+func (user *UserBase) GetSetting() dto.UserSetting {
+	setting := dto.UserSetting{}
+	if user.Setting != "" {
+		err := common.Unmarshal([]byte(user.Setting), &setting)
+		if err != nil {
+			common.SysError("failed to unmarshal setting: " + err.Error())
+		}
 	}
-	return common.StrToMap(user.Setting)
-}
-
-func (user *UserBase) SetSetting(setting map[string]interface{}) {
-	settingBytes, err := json.Marshal(setting)
-	if err != nil {
-		common.SysError("failed to marshal setting: " + err.Error())
-		return
-	}
-	user.Setting = string(settingBytes)
+	return setting
 }
 
 // getUserCacheKey returns the key for user cache
@@ -70,7 +65,7 @@ func updateUserCache(user User) error {
 	return common.RedisHSetObj(
 		getUserCacheKey(user.Id),
 		user.ToBaseUser(),
-		time.Duration(constant.RedisKeyCacheSeconds())*time.Second,
+		time.Duration(common.RedisKeyCacheSeconds())*time.Second,
 	)
 }
 
@@ -174,11 +169,10 @@ func getUserNameCache(userId int) (string, error) {
 	return cache.Username, nil
 }
 
-func getUserSettingCache(userId int) (map[string]interface{}, error) {
-	setting := make(map[string]interface{})
+func getUserSettingCache(userId int) (dto.UserSetting, error) {
 	cache, err := GetUserCache(userId)
 	if err != nil {
-		return setting, err
+		return dto.UserSetting{}, err
 	}
 	return cache.GetSetting(), nil
 }

relay/audio_handler.go

@@ -3,7 +3,6 @@ package relay
 import (
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
@@ -12,7 +11,10 @@ import (
 	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
+	"one-api/types"
 	"strings"
+
+	"github.com/gin-gonic/gin"
 )
 
 func getAndValidAudioRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.AudioRequest, error) {
@@ -54,13 +56,13 @@ func getAndValidAudioRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.
 	return audioRequest, nil
 }
 
-func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
+func AudioHelper(c *gin.Context) (newAPIError *types.NewAPIError) {
 	relayInfo := relaycommon.GenRelayInfoOpenAIAudio(c)
 	audioRequest, err := getAndValidAudioRequest(c, relayInfo)
 
 	if err != nil {
 		common.LogError(c, fmt.Sprintf("getAndValidAudioRequest failed: %s", err.Error()))
-		return service.OpenAIErrorWrapper(err, "invalid_audio_request", http.StatusBadRequest)
+		return types.NewError(err, types.ErrorCodeInvalidRequest)
 	}
 
 	promptTokens := 0
@@ -73,7 +75,7 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 
 	priceData, err := helper.ModelPriceHelper(c, relayInfo, preConsumedTokens, 0)
 	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "model_price_error", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeModelPriceError)
 	}
 
 	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
@@ -88,23 +90,23 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 
 	err = helper.ModelMappedHelper(c, relayInfo, audioRequest)
 	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeChannelModelMappedError)
 	}
 
 	adaptor := GetAdaptor(relayInfo.ApiType)
 	if adaptor == nil {
-		return service.OpenAIErrorWrapperLocal(fmt.Errorf("invalid api type: %d", relayInfo.ApiType), "invalid_api_type", http.StatusBadRequest)
+		return types.NewError(fmt.Errorf("invalid api type: %d", relayInfo.ApiType), types.ErrorCodeInvalidApiType)
 	}
 	adaptor.Init(relayInfo)
 
 	ioReader, err := adaptor.ConvertAudioRequest(c, relayInfo, *audioRequest)
 	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "convert_request_failed", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeConvertRequestFailed)
 	}
 
 	resp, err := adaptor.DoRequest(c, relayInfo, ioReader)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "do_request_failed", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeDoRequestFailed)
 	}
 	statusCodeMappingStr := c.GetString("status_code_mapping")
 
@@ -112,18 +114,18 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	if resp != nil {
 		httpResp = resp.(*http.Response)
 		if httpResp.StatusCode != http.StatusOK {
-			openaiErr = service.RelayErrorHandler(httpResp, false)
+			newAPIError = service.RelayErrorHandler(httpResp, false)
 			// reset status code 重置状态码
-			service.ResetStatusCode(openaiErr, statusCodeMappingStr)
-			return openaiErr
+			service.ResetStatusCode(newAPIError, statusCodeMappingStr)
+			return newAPIError
 		}
 	}
 
-	usage, openaiErr := adaptor.DoResponse(c, httpResp, relayInfo)
-	if openaiErr != nil {
+	usage, newAPIError := adaptor.DoResponse(c, httpResp, relayInfo)
+	if newAPIError != nil {
 		// reset status code 重置状态码
-		service.ResetStatusCode(openaiErr, statusCodeMappingStr)
-		return openaiErr
+		service.ResetStatusCode(newAPIError, statusCodeMappingStr)
+		return newAPIError
 	}
 
 	postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")

relay/channel/adapter.go

@@ -5,6 +5,7 @@ import (
 	"net/http"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -21,7 +22,7 @@ type Adaptor interface {
 	ConvertImageRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.ImageRequest) (any, error)
 	ConvertOpenAIResponsesRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.OpenAIResponsesRequest) (any, error)
 	DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (any, error)
-	DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode)
+	DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError)
 	GetModelList() []string
 	GetChannelName() string
 	ConvertClaudeRequest(c *gin.Context, info *relaycommon.RelayInfo, request *dto.ClaudeRequest) (any, error)
@@ -45,5 +46,5 @@ type TaskAdaptor interface {
 	// FetchTask
 	FetchTask(baseUrl, key string, body map[string]any) (*http.Response, error)
 
-	ParseResultUrl(resp map[string]any) (string, error)
+	ParseTaskResult(respBody []byte) (*relaycommon.TaskInfo, error)
 }

relay/channel/ali/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -30,7 +31,7 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 	var fullRequestURL string
 	switch info.RelayMode {
 	case constant.RelayModeEmbeddings:
-		fullRequestURL = fmt.Sprintf("%s/api/v1/services/embeddings/text-embedding/text-embedding", info.BaseUrl)
+		fullRequestURL = fmt.Sprintf("%s/compatible-mode/v1/embeddings", info.BaseUrl)
 	case constant.RelayModeRerank:
 		fullRequestURL = fmt.Sprintf("%s/api/v1/services/rerank/text-rerank/text-rerank", info.BaseUrl)
 	case constant.RelayModeImagesGenerations:
@@ -82,7 +83,7 @@ func (a *Adaptor) ConvertRerankRequest(c *gin.Context, relayMode int, request dt
 }
 
 func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.EmbeddingRequest) (any, error) {
-	return embeddingRequestOpenAI2Ali(request), nil
+	return request, nil
 }
 
 func (a *Adaptor) ConvertAudioRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.AudioRequest) (io.Reader, error) {
@@ -99,7 +100,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayMode {
 	case constant.RelayModeImagesGenerations:
 		err, usage = aliImageHandler(c, resp, info)
@@ -109,9 +110,9 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 		err, usage = RerankHandler(c, resp, info)
 	default:
 		if info.IsStream {
-			err, usage = openai.OaiStreamHandler(c, resp, info)
+			usage, err = openai.OaiStreamHandler(c, info, resp)
 		} else {
-			err, usage = openai.OpenaiHandler(c, resp, info)
+			usage, err = openai.OpenaiHandler(c, info, resp)
 		}
 	}
 	return

relay/channel/ali/image.go

@@ -4,15 +4,17 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
 )
 
 func oaiImage2Ali(request dto.ImageRequest) *AliImageRequest {
@@ -124,52 +126,46 @@ func responseAli2OpenAIImage(c *gin.Context, response *AliResponse, info *relayc
 	return &imageResponse
 }
 
-func aliImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func aliImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*types.NewAPIError, *dto.Usage) {
 	responseFormat := c.GetString("response_format")
 
 	var aliTaskResponse AliResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeReadResponseBodyFailed), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &aliTaskResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 
 	if aliTaskResponse.Message != "" {
 		common.LogError(c, "ali_async_task_failed: "+aliTaskResponse.Message)
-		return service.OpenAIErrorWrapper(errors.New(aliTaskResponse.Message), "ali_async_task_failed", http.StatusInternalServerError), nil
+		return types.NewError(errors.New(aliTaskResponse.Message), types.ErrorCodeBadResponse), nil
 	}
 
 	aliResponse, _, err := asyncTaskWait(info, aliTaskResponse.Output.TaskId)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "ali_async_task_wait_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponse), nil
 	}
 
 	if aliResponse.Output.TaskStatus != "SUCCEEDED" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: aliResponse.Output.Message,
-				Type:    "ali_error",
-				Param:   "",
-				Code:    aliResponse.Output.Code,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return types.WithOpenAIError(types.OpenAIError{
+			Message: aliResponse.Output.Message,
+			Type:    "ali_error",
+			Param:   "",
+			Code:    aliResponse.Output.Code,
+		}, resp.StatusCode), nil
 	}
 
 	fullTextResponse := responseAli2OpenAIImage(c, aliResponse, info, responseFormat)
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	return nil, nil
+	c.Writer.Write(jsonResponse)
+	return nil, &dto.Usage{}
 }

relay/channel/ali/rerank.go

@@ -4,9 +4,10 @@ import (
 	"encoding/json"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
-	"one-api/service"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -30,32 +31,26 @@ func ConvertRerankRequest(request dto.RerankRequest) *AliRerankRequest {
 	}
 }
 
-func RerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func RerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*types.NewAPIError, *dto.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeReadResponseBodyFailed), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 
 	var aliResponse AliRerankResponse
 	err = json.Unmarshal(responseBody, &aliResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 
 	if aliResponse.Code != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: aliResponse.Message,
-				Type:    aliResponse.Code,
-				Param:   aliResponse.RequestId,
-				Code:    aliResponse.Code,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return types.WithOpenAIError(types.OpenAIError{
+			Message: aliResponse.Message,
+			Type:    aliResponse.Code,
+			Param:   aliResponse.RequestId,
+			Code:    aliResponse.Code,
+		}, resp.StatusCode), nil
 	}
 
 	usage := dto.Usage{
@@ -70,14 +65,10 @@ func RerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayI
 
 	jsonResponse, err := json.Marshal(rerankResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "write_response_body_failed", http.StatusInternalServerError), nil
-	}
-
+	c.Writer.Write(jsonResponse)
 	return nil, &usage
 }

relay/channel/ali/text.go

@@ -8,9 +8,10 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	"one-api/relay/helper"
-	"one-api/service"
 	"strings"
 
+	"one-api/types"
+
 	"github.com/gin-gonic/gin"
 )
 
@@ -38,42 +39,26 @@ func embeddingRequestOpenAI2Ali(request dto.EmbeddingRequest) *AliEmbeddingReque
 	}
 }
 
-func aliEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
-	var aliResponse AliEmbeddingResponse
-	err := json.NewDecoder(resp.Body).Decode(&aliResponse)
+func aliEmbeddingHandler(c *gin.Context, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
+	var fullTextResponse dto.OpenAIEmbeddingResponse
+	err := json.NewDecoder(resp.Body).Decode(&fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-
-	if aliResponse.Code != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: aliResponse.Message,
-				Type:    aliResponse.Code,
-				Param:   aliResponse.RequestId,
-				Code:    aliResponse.Code,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
-	}
+	common.CloseResponseBodyGracefully(resp)
 
 	model := c.GetString("model")
 	if model == "" {
 		model = "text-embedding-v4"
 	}
-	fullTextResponse := embeddingResponseAli2OpenAI(&aliResponse, model)
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
+	c.Writer.Write(jsonResponse)
 	return nil, &fullTextResponse.Usage
 }
 
@@ -135,7 +120,7 @@ func streamResponseAli2OpenAI(aliResponse *AliResponse) *dto.ChatCompletionsStre
 	return &response
 }
 
-func aliStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func aliStreamHandler(c *gin.Context, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	var usage dto.Usage
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
@@ -186,42 +171,33 @@ func aliStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWith
 			return false
 		}
 	})
-	err := resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
+	common.CloseResponseBodyGracefully(resp)
 	return nil, &usage
 }
 
-func aliHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func aliHandler(c *gin.Context, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	var aliResponse AliResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeReadResponseBodyFailed), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &aliResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	if aliResponse.Code != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: aliResponse.Message,
-				Type:    aliResponse.Code,
-				Param:   aliResponse.RequestId,
-				Code:    aliResponse.Code,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return types.WithOpenAIError(types.OpenAIError{
+			Message: aliResponse.Message,
+			Type:    "ali_error",
+			Param:   aliResponse.RequestId,
+			Code:    aliResponse.Code,
+		}, resp.StatusCode), nil
 	}
 	fullTextResponse := responseAli2OpenAI(&aliResponse)
-	jsonResponse, err := json.Marshal(fullTextResponse)
+	jsonResponse, err := common.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)

relay/channel/api_request.go

@@ -206,8 +206,8 @@ func sendPingData(c *gin.Context, mutex *sync.Mutex) error {
 func doRequest(c *gin.Context, req *http.Request, info *common.RelayInfo) (*http.Response, error) {
 	var client *http.Client
 	var err error
-	if proxyURL, ok := info.ChannelSetting["proxy"]; ok {
-		client, err = service.NewProxyHttpClient(proxyURL.(string))
+	if info.ChannelSetting.Proxy != "" {
+		client, err = service.NewProxyHttpClient(info.ChannelSetting.Proxy)
 		if err != nil {
 			return nil, fmt.Errorf("new proxy http client failed: %w", err)
 		}

relay/channel/aws/adaptor.go

@@ -8,6 +8,7 @@ import (
 	"one-api/relay/channel/claude"
 	relaycommon "one-api/relay/common"
 	"one-api/setting/model_setting"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -84,7 +85,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return nil, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
 		err, usage = awsStreamHandler(c, resp, info, a.RequestMode)
 	} else {

relay/channel/aws/relay-aws.go

@@ -3,19 +3,22 @@ package aws
 import (
 	"encoding/json"
 	"fmt"
-	"github.com/gin-gonic/gin"
-	"github.com/pkg/errors"
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
 	"one-api/relay/channel/claude"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
+	"one-api/types"
 	"strings"
 
+	"github.com/gin-gonic/gin"
+	"github.com/pkg/errors"
+
 	"github.com/aws/aws-sdk-go-v2/aws"
 	"github.com/aws/aws-sdk-go-v2/credentials"
 	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime"
-	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime/types"
+	bedrockruntimeTypes "github.com/aws/aws-sdk-go-v2/service/bedrockruntime/types"
 )
 
 func newAwsClient(c *gin.Context, info *relaycommon.RelayInfo) (*bedrockruntime.Client, error) {
@@ -65,24 +68,21 @@ func awsModelCrossRegion(awsModelId, awsRegionPrefix string) string {
 	return modelPrefix + "." + awsModelId
 }
 
-func awsModelID(requestModel string) (string, error) {
+func awsModelID(requestModel string) string {
 	if awsModelID, ok := awsModelIDMap[requestModel]; ok {
-		return awsModelID, nil
+		return awsModelID
 	}
 
-	return requestModel, nil
+	return requestModel
 }
 
-func awsHandler(c *gin.Context, info *relaycommon.RelayInfo, requestMode int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func awsHandler(c *gin.Context, info *relaycommon.RelayInfo, requestMode int) (*types.NewAPIError, *dto.Usage) {
 	awsCli, err := newAwsClient(c, info)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
+		return types.NewError(err, types.ErrorCodeChannelAwsClientError), nil
 	}
 
-	awsModelId, err := awsModelID(c.GetString("request_model"))
-	if err != nil {
-		return wrapErr(errors.Wrap(err, "awsModelID")), nil
-	}
+	awsModelId := awsModelID(c.GetString("request_model"))
 
 	awsRegionPrefix := awsRegionPrefix(awsCli.Options().Region)
 	canCrossRegion := awsModelCanCrossRegion(awsModelId, awsRegionPrefix)
@@ -98,42 +98,42 @@ func awsHandler(c *gin.Context, info *relaycommon.RelayInfo, requestMode int) (*
 
 	claudeReq_, ok := c.Get("converted_request")
 	if !ok {
-		return wrapErr(errors.New("request not found")), nil
+		return types.NewError(errors.New("aws claude request not found"), types.ErrorCodeInvalidRequest), nil
 	}
 	claudeReq := claudeReq_.(*dto.ClaudeRequest)
 	awsClaudeReq := copyRequest(claudeReq)
 	awsReq.Body, err = json.Marshal(awsClaudeReq)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "marshal request")), nil
+		return types.NewError(errors.Wrap(err, "marshal request"), types.ErrorCodeBadResponseBody), nil
 	}
 
 	awsResp, err := awsCli.InvokeModel(c.Request.Context(), awsReq)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "InvokeModel")), nil
+		return types.NewError(errors.Wrap(err, "InvokeModel"), types.ErrorCodeChannelAwsClientError), nil
 	}
 
 	claudeInfo := &claude.ClaudeResponseInfo{
-		ResponseId:   fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
+		ResponseId:   helper.GetResponseID(c),
 		Created:      common.GetTimestamp(),
 		Model:        info.UpstreamModelName,
 		ResponseText: strings.Builder{},
 		Usage:        &dto.Usage{},
 	}
 
-	claude.HandleClaudeResponseData(c, info, claudeInfo, awsResp.Body, RequestModeMessage)
+	handlerErr := claude.HandleClaudeResponseData(c, info, claudeInfo, awsResp.Body, RequestModeMessage)
+	if handlerErr != nil {
+		return handlerErr, nil
+	}
 	return nil, claudeInfo.Usage
 }
 
-func awsStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, requestMode int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func awsStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, requestMode int) (*types.NewAPIError, *dto.Usage) {
 	awsCli, err := newAwsClient(c, info)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
+		return types.NewError(err, types.ErrorCodeChannelAwsClientError), nil
 	}
 
-	awsModelId, err := awsModelID(c.GetString("request_model"))
-	if err != nil {
-		return wrapErr(errors.Wrap(err, "awsModelID")), nil
-	}
+	awsModelId := awsModelID(c.GetString("request_model"))
 
 	awsRegionPrefix := awsRegionPrefix(awsCli.Options().Region)
 	canCrossRegion := awsModelCanCrossRegion(awsModelId, awsRegionPrefix)
@@ -149,25 +149,25 @@ func awsStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 
 	claudeReq_, ok := c.Get("converted_request")
 	if !ok {
-		return wrapErr(errors.New("request not found")), nil
+		return types.NewError(errors.New("aws claude request not found"), types.ErrorCodeInvalidRequest), nil
 	}
 	claudeReq := claudeReq_.(*dto.ClaudeRequest)
 
 	awsClaudeReq := copyRequest(claudeReq)
 	awsReq.Body, err = json.Marshal(awsClaudeReq)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "marshal request")), nil
+		return types.NewError(errors.Wrap(err, "marshal request"), types.ErrorCodeBadResponseBody), nil
 	}
 
 	awsResp, err := awsCli.InvokeModelWithResponseStream(c.Request.Context(), awsReq)
 	if err != nil {
-		return wrapErr(errors.Wrap(err, "InvokeModelWithResponseStream")), nil
+		return types.NewError(errors.Wrap(err, "InvokeModelWithResponseStream"), types.ErrorCodeChannelAwsClientError), nil
 	}
 	stream := awsResp.GetStream()
 	defer stream.Close()
 
 	claudeInfo := &claude.ClaudeResponseInfo{
-		ResponseId:   fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
+		ResponseId:   helper.GetResponseID(c),
 		Created:      common.GetTimestamp(),
 		Model:        info.UpstreamModelName,
 		ResponseText: strings.Builder{},
@@ -176,18 +176,18 @@ func awsStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 
 	for event := range stream.Events() {
 		switch v := event.(type) {
-		case *types.ResponseStreamMemberChunk:
+		case *bedrockruntimeTypes.ResponseStreamMemberChunk:
 			info.SetFirstResponseTime()
 			respErr := claude.HandleStreamResponseData(c, info, claudeInfo, string(v.Value.Bytes), RequestModeMessage)
 			if respErr != nil {
 				return respErr, nil
 			}
-		case *types.UnknownUnionMember:
+		case *bedrockruntimeTypes.UnknownUnionMember:
 			fmt.Println("unknown tag:", v.Tag)
-			return wrapErr(errors.New("unknown response type")), nil
+			return types.NewError(errors.New("unknown response type"), types.ErrorCodeInvalidRequest), nil
 		default:
 			fmt.Println("union is nil or unknown type")
-			return wrapErr(errors.New("nil or unknown response type")), nil
+			return types.NewError(errors.New("nil or unknown response type"), types.ErrorCodeInvalidRequest), nil
 		}
 	}
 

relay/channel/baidu/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -140,15 +141,15 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = baiduStreamHandler(c, resp)
+		err, usage = baiduStreamHandler(c, info, resp)
 	} else {
 		switch info.RelayMode {
 		case constant.RelayModeEmbeddings:
-			err, usage = baiduEmbeddingHandler(c, resp)
+			err, usage = baiduEmbeddingHandler(c, info, resp)
 		default:
-			err, usage = baiduHandler(c, resp)
+			err, usage = baiduHandler(c, info, resp)
 		}
 	}
 	return

relay/channel/baidu/relay-baidu.go

@@ -1,21 +1,23 @@
 package baidu
 
 import (
-	"bufio"
 	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 	"sync"
 	"time"
+
+	"github.com/gin-gonic/gin"
 )
 
 // https://cloud.baidu.com/doc/WENXINWORKSHOP/s/flfmc9do2
@@ -110,98 +112,49 @@ func embeddingResponseBaidu2OpenAI(response *BaiduEmbeddingResponse) *dto.OpenAI
 	return &openAIEmbeddingResponse
 }
 
-func baiduStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
-	var usage dto.Usage
-	scanner := bufio.NewScanner(resp.Body)
-	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
-		if atEOF && len(data) == 0 {
-			return 0, nil, nil
-		}
-		if i := strings.Index(string(data), "\n"); i >= 0 {
-			return i + 1, data[0:i], nil
-		}
-		if atEOF {
-			return len(data), data, nil
-		}
-		return 0, nil, nil
-	})
-	dataChan := make(chan string)
-	stopChan := make(chan bool)
-	go func() {
-		for scanner.Scan() {
-			data := scanner.Text()
-			if len(data) < 6 { // ignore blank line or wrong format
-				continue
-			}
-			data = data[6:]
-			dataChan <- data
-		}
-		stopChan <- true
-	}()
-	helper.SetEventStreamHeaders(c)
-	c.Stream(func(w io.Writer) bool {
-		select {
-		case data := <-dataChan:
-			var baiduResponse BaiduChatStreamResponse
-			err := json.Unmarshal([]byte(data), &baiduResponse)
-			if err != nil {
-				common.SysError("error unmarshalling stream response: " + err.Error())
-				return true
-			}
-			if baiduResponse.Usage.TotalTokens != 0 {
-				usage.TotalTokens = baiduResponse.Usage.TotalTokens
-				usage.PromptTokens = baiduResponse.Usage.PromptTokens
-				usage.CompletionTokens = baiduResponse.Usage.TotalTokens - baiduResponse.Usage.PromptTokens
-			}
-			response := streamResponseBaidu2OpenAI(&baiduResponse)
-			jsonResponse, err := json.Marshal(response)
-			if err != nil {
-				common.SysError("error marshalling stream response: " + err.Error())
-				return true
-			}
-			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonResponse)})
+func baiduStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
+	usage := &dto.Usage{}
+	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
+		var baiduResponse BaiduChatStreamResponse
+		err := common.Unmarshal([]byte(data), &baiduResponse)
+		if err != nil {
+			common.SysError("error unmarshalling stream response: " + err.Error())
 			return true
-		case <-stopChan:
-			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
-			return false
 		}
+		if baiduResponse.Usage.TotalTokens != 0 {
+			usage.TotalTokens = baiduResponse.Usage.TotalTokens
+			usage.PromptTokens = baiduResponse.Usage.PromptTokens
+			usage.CompletionTokens = baiduResponse.Usage.TotalTokens - baiduResponse.Usage.PromptTokens
+		}
+		response := streamResponseBaidu2OpenAI(&baiduResponse)
+		err = helper.ObjectData(c, response)
+		if err != nil {
+			common.SysError("error sending stream response: " + err.Error())
+		}
+		return true
 	})
-	err := resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	return nil, &usage
+	common.CloseResponseBodyGracefully(resp)
+	return nil, usage
 }
 
-func baiduHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func baiduHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	var baiduResponse BaiduChatResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	if baiduResponse.ErrorMsg != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: baiduResponse.ErrorMsg,
-				Type:    "baidu_error",
-				Param:   "",
-				Code:    baiduResponse.ErrorCode,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return types.NewError(fmt.Errorf(baiduResponse.ErrorMsg), types.ErrorCodeBadResponseBody), nil
 	}
 	fullTextResponse := responseBaidu2OpenAI(&baiduResponse)
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
@@ -209,35 +162,24 @@ func baiduHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStat
 	return nil, &fullTextResponse.Usage
 }
 
-func baiduEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func baiduEmbeddingHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	var baiduResponse BaiduEmbeddingResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	if baiduResponse.ErrorMsg != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: baiduResponse.ErrorMsg,
-				Type:    "baidu_error",
-				Param:   "",
-				Code:    baiduResponse.ErrorCode,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return types.NewError(fmt.Errorf(baiduResponse.ErrorMsg), types.ErrorCodeBadResponseBody), nil
 	}
 	fullTextResponse := embeddingResponseBaidu2OpenAI(&baiduResponse)
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
@@ -280,7 +222,7 @@ func getBaiduAccessTokenHelper(apiKey string) (*BaiduAccessToken, error) {
 	}
 	req.Header.Add("Content-Type", "application/json")
 	req.Header.Add("Accept", "application/json")
-	res, err := service.GetImpatientHttpClient().Do(req)
+	res, err := service.GetHttpClient().Do(req)
 	if err != nil {
 		return nil, err
 	}

relay/channel/baidu_v2/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -42,7 +43,16 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Header, info *relaycommon.RelayInfo) error {
 	channel.SetupApiRequestHeader(info, c, req)
-	req.Set("Authorization", "Bearer "+info.ApiKey)
+        keyParts := strings.Split(info.ApiKey, "|")
+	if len(keyParts) == 0 || keyParts[0] == "" {
+        	return errors.New("invalid API key: authorization token is required")
+        }
+	 if len(keyParts) > 1 {
+               if keyParts[1] != "" {
+                       req.Set("appid", keyParts[1])
+               }
+        }
+	req.Set("Authorization", "Bearer "+keyParts[0])
 	return nil
 }
 
@@ -83,11 +93,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = openai.OaiStreamHandler(c, resp, info)
+		usage, err = openai.OaiStreamHandler(c, info, resp)
 	} else {
-		err, usage = openai.OpenaiHandler(c, resp, info)
+		usage, err = openai.OpenaiHandler(c, info, resp)
 	}
 	return
 }

relay/channel/claude/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/setting/model_setting"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -94,7 +95,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
 		err, usage = ClaudeStreamHandler(c, resp, info, a.RequestMode)
 	} else {

relay/channel/claude/relay-claude.go

@@ -12,6 +12,7 @@ import (
 	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting/model_setting"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -125,7 +126,7 @@ func RequestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*dto.Cla
 
 	if textRequest.Reasoning != nil {
 		var reasoning openrouter.RequestReasoning
-		if err := common.DecodeJson(textRequest.Reasoning, &reasoning); err != nil {
+		if err := common.Unmarshal(textRequest.Reasoning, &reasoning); err != nil {
 			return nil, err
 		}
 
@@ -517,22 +518,15 @@ func FormatClaudeResponseInfo(requestMode int, claudeResponse *dto.ClaudeRespons
 	return true
 }
 
-func HandleStreamResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data string, requestMode int) *dto.OpenAIErrorWithStatusCode {
+func HandleStreamResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data string, requestMode int) *types.NewAPIError {
 	var claudeResponse dto.ClaudeResponse
-	err := common.DecodeJsonStr(data, &claudeResponse)
+	err := common.UnmarshalJsonStr(data, &claudeResponse)
 	if err != nil {
 		common.SysError("error unmarshalling stream response: " + err.Error())
-		return service.OpenAIErrorWrapper(err, "stream_response_error", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if claudeResponse.Error != nil && claudeResponse.Error.Type != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Code:    "stream_response_error",
-				Type:    claudeResponse.Error.Type,
-				Message: claudeResponse.Error.Message,
-			},
-			StatusCode: http.StatusInternalServerError,
-		}
+		return types.WithClaudeError(*claudeResponse.Error, http.StatusInternalServerError)
 	}
 	if info.RelayFormat == relaycommon.RelayFormatClaude {
 		FormatClaudeResponseInfo(requestMode, &claudeResponse, nil, claudeInfo)
@@ -593,15 +587,15 @@ func HandleStreamFinalResponse(c *gin.Context, info *relaycommon.RelayInfo, clau
 	}
 }
 
-func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, requestMode int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, requestMode int) (*types.NewAPIError, *dto.Usage) {
 	claudeInfo := &ClaudeResponseInfo{
-		ResponseId:   fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
+		ResponseId:   helper.GetResponseID(c),
 		Created:      common.GetTimestamp(),
 		Model:        info.UpstreamModelName,
 		ResponseText: strings.Builder{},
 		Usage:        &dto.Usage{},
 	}
-	var err *dto.OpenAIErrorWithStatusCode
+	var err *types.NewAPIError
 	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
 		err = HandleStreamResponseData(c, info, claudeInfo, data, requestMode)
 		if err != nil {
@@ -617,21 +611,14 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 	return nil, claudeInfo.Usage
 }
 
-func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data []byte, requestMode int) *dto.OpenAIErrorWithStatusCode {
+func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data []byte, requestMode int) *types.NewAPIError {
 	var claudeResponse dto.ClaudeResponse
-	err := common.DecodeJson(data, &claudeResponse)
+	err := common.Unmarshal(data, &claudeResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_claude_response_failed", http.StatusInternalServerError)
+		return types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if claudeResponse.Error != nil && claudeResponse.Error.Type != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: claudeResponse.Error.Message,
-				Type:    claudeResponse.Error.Type,
-				Code:    claudeResponse.Error.Type,
-			},
-			StatusCode: http.StatusInternalServerError,
-		}
+		return types.WithClaudeError(*claudeResponse.Error, http.StatusInternalServerError)
 	}
 	if requestMode == RequestModeCompletion {
 		completionTokens := service.CountTextToken(claudeResponse.Completion, info.OriginModelName)
@@ -652,20 +639,21 @@ func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claud
 		openaiResponse.Usage = *claudeInfo.Usage
 		responseData, err = json.Marshal(openaiResponse)
 		if err != nil {
-			return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError)
+			return types.NewError(err, types.ErrorCodeBadResponseBody)
 		}
 	case relaycommon.RelayFormatClaude:
 		responseData = data
 	}
-	c.Writer.Header().Set("Content-Type", "application/json")
-	c.Writer.WriteHeader(http.StatusOK)
-	_, err = c.Writer.Write(responseData)
+
+	common.IOCopyBytesGracefully(c, nil, responseData)
 	return nil
 }
 
-func ClaudeHandler(c *gin.Context, resp *http.Response, requestMode int, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func ClaudeHandler(c *gin.Context, resp *http.Response, requestMode int, info *relaycommon.RelayInfo) (*types.NewAPIError, *dto.Usage) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	claudeInfo := &ClaudeResponseInfo{
-		ResponseId:   fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
+		ResponseId:   helper.GetResponseID(c),
 		Created:      common.GetTimestamp(),
 		Model:        info.UpstreamModelName,
 		ResponseText: strings.Builder{},
@@ -673,9 +661,8 @@ func ClaudeHandler(c *gin.Context, resp *http.Response, requestMode int, info *r
 	}
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
-	resp.Body.Close()
 	if common.DebugEnabled {
 		println("responseBody: ", string(responseBody))
 	}

relay/channel/cloudflare/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -94,20 +95,20 @@ func (a *Adaptor) ConvertImageRequest(c *gin.Context, info *relaycommon.RelayInf
 	return nil, errors.New("not implemented")
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayMode {
 	case constant.RelayModeEmbeddings:
 		fallthrough
 	case constant.RelayModeChatCompletions:
 		if info.IsStream {
-			err, usage = cfStreamHandler(c, resp, info)
+			err, usage = cfStreamHandler(c, info, resp)
 		} else {
-			err, usage = cfHandler(c, resp, info)
+			err, usage = cfHandler(c, info, resp)
 		}
 	case constant.RelayModeAudioTranslation:
 		fallthrough
 	case constant.RelayModeAudioTranscription:
-		err, usage = cfSTTHandler(c, resp, info)
+		err, usage = cfSTTHandler(c, info, resp)
 	}
 	return
 }

relay/channel/cloudflare/relay_cloudflare.go

@@ -3,7 +3,6 @@ package cloudflare
 import (
 	"bufio"
 	"encoding/json"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -11,8 +10,11 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
 )
 
 func convertCf2CompletionsRequest(textRequest dto.GeneralOpenAIRequest) *CfRequest {
@@ -25,7 +27,7 @@ func convertCf2CompletionsRequest(textRequest dto.GeneralOpenAIRequest) *CfReque
 	}
 }
 
-func cfStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cfStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 
@@ -81,27 +83,21 @@ func cfStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	}
 	helper.Done(c)
 
-	err := resp.Body.Close()
-	if err != nil {
-		common.LogError(c, "close_response_body_failed: "+err.Error())
-	}
+	common.CloseResponseBodyGracefully(resp)
 
 	return nil, usage
 }
 
-func cfHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cfHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	var response dto.TextResponse
 	err = json.Unmarshal(responseBody, &response)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	response.Model = info.UpstreamModelName
 	var responseText string
@@ -113,7 +109,7 @@ func cfHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo)
 	response.Id = helper.GetResponseID(c)
 	jsonResponse, err := json.Marshal(response)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
@@ -121,19 +117,16 @@ func cfHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo)
 	return nil, usage
 }
 
-func cfSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cfSTTHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	var cfResp CfAudioResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &cfResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 
 	audioResp := &dto.AudioResponse{
@@ -142,7 +135,7 @@ func cfSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayIn
 
 	jsonResponse, err := json.Marshal(audioResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)

relay/channel/cohere/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -71,14 +72,14 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 	return nil, errors.New("not implemented")
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.RelayMode == constant.RelayModeRerank {
 		err, usage = cohereRerankHandler(c, resp, info)
 	} else {
 		if info.IsStream {
-			err, usage = cohereStreamHandler(c, resp, info)
+			err, usage = cohereStreamHandler(c, info, resp)
 		} else {
-			err, usage = cohereHandler(c, resp, info.UpstreamModelName, info.PromptTokens)
+			err, usage = cohereHandler(c, info, resp)
 		}
 	}
 	return

relay/channel/cohere/relay-cohere.go

@@ -3,7 +3,6 @@ package cohere
 import (
 	"bufio"
 	"encoding/json"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -11,8 +10,11 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
 )
 
 func requestOpenAI2Cohere(textRequest dto.GeneralOpenAIRequest) *CohereRequest {
@@ -76,7 +78,7 @@ func stopReasonCohere2OpenAI(reason string) string {
 	}
 }
 
-func cohereStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cohereStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	responseId := helper.GetResponseID(c)
 	createdTime := common.GetTimestamp()
 	usage := &dto.Usage{}
@@ -167,20 +169,17 @@ func cohereStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 	return nil, usage
 }
 
-func cohereHandler(c *gin.Context, resp *http.Response, modelName string, promptTokens int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cohereHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	createdTime := common.GetTimestamp()
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	var cohereResp CohereResponseResult
 	err = json.Unmarshal(responseBody, &cohereResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	usage := dto.Usage{}
 	usage.PromptTokens = cohereResp.Meta.BilledUnits.InputTokens
@@ -191,7 +190,7 @@ func cohereHandler(c *gin.Context, resp *http.Response, modelName string, prompt
 	openaiResp.Id = cohereResp.ResponseId
 	openaiResp.Created = createdTime
 	openaiResp.Object = "chat.completion"
-	openaiResp.Model = modelName
+	openaiResp.Model = info.UpstreamModelName
 	openaiResp.Usage = usage
 
 	openaiResp.Choices = []dto.OpenAITextResponseChoice{
@@ -204,7 +203,7 @@ func cohereHandler(c *gin.Context, resp *http.Response, modelName string, prompt
 
 	jsonResponse, err := json.Marshal(openaiResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
@@ -212,19 +211,16 @@ func cohereHandler(c *gin.Context, resp *http.Response, modelName string, prompt
 	return nil, &usage
 }
 
-func cohereRerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cohereRerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*types.NewAPIError, *dto.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	var cohereResp CohereRerankResponseResult
 	err = json.Unmarshal(responseBody, &cohereResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	usage := dto.Usage{}
 	if cohereResp.Meta.BilledUnits.InputTokens == 0 {
@@ -243,7 +239,7 @@ func cohereRerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 
 	jsonResponse, err := json.Marshal(rerankResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)

relay/channel/coze/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/dto"
 	"one-api/relay/channel"
 	"one-api/relay/common"
+	"one-api/types"
 	"time"
 
 	"github.com/gin-gonic/gin"
@@ -95,11 +96,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *common.RelayInfo, requestBody
 }
 
 // DoResponse implements channel.Adaptor.
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *common.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *common.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = cozeChatStreamHandler(c, resp, info)
+		err, usage = cozeChatStreamHandler(c, info, resp)
 	} else {
-		err, usage = cozeChatHandler(c, resp, info)
+		err, usage = cozeChatHandler(c, info, resp)
 	}
 	return
 }

relay/channel/coze/relay-coze.go

@@ -12,6 +12,7 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -43,25 +44,22 @@ func convertCozeChatRequest(c *gin.Context, request dto.GeneralOpenAIRequest) *C
 	return cozeRequest
 }
 
-func cozeChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cozeChatHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
+	common.CloseResponseBodyGracefully(resp)
 	// convert coze response to openai response
 	var response dto.TextResponse
 	var cozeResponse CozeChatDetailResponse
 	response.Model = info.UpstreamModelName
 	err = json.Unmarshal(responseBody, &cozeResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	if cozeResponse.Code != 0 {
-		return service.OpenAIErrorWrapper(errors.New(cozeResponse.Msg), fmt.Sprintf("%d", cozeResponse.Code), http.StatusInternalServerError), nil
+		return types.NewError(errors.New(cozeResponse.Msg), types.ErrorCodeBadResponseBody), nil
 	}
 	// 从上下文获取 usage
 	var usage dto.Usage
@@ -88,7 +86,7 @@ func cozeChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	}
 	jsonResponse, err := json.Marshal(response)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
@@ -97,7 +95,7 @@ func cozeChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	return nil, &usage
 }
 
-func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func cozeChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*types.NewAPIError, *dto.Usage) {
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 	helper.SetEventStreamHeaders(c)
@@ -138,7 +136,7 @@ func cozeChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommo
 	}
 
 	if err := scanner.Err(); err != nil {
-		return service.OpenAIErrorWrapper(err, "stream_scanner_error", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeBadResponseBody), nil
 	}
 	helper.Done(c)
 
@@ -281,8 +279,8 @@ func getChatDetail(a *Adaptor, c *gin.Context, info *relaycommon.RelayInfo) (*ht
 func doRequest(req *http.Request, info *relaycommon.RelayInfo) (*http.Response, error) {
 	var client *http.Client
 	var err error // 声明 err 变量
-	if proxyURL, ok := info.ChannelSetting["proxy"]; ok {
-		client, err = service.NewProxyHttpClient(proxyURL.(string))
+	if info.ChannelSetting.Proxy != "" {
+		client, err = service.NewProxyHttpClient(info.ChannelSetting.Proxy)
 		if err != nil {
 			return nil, fmt.Errorf("new proxy http client failed: %w", err)
 		}

relay/channel/deepseek/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -81,11 +82,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = openai.OaiStreamHandler(c, resp, info)
+		usage, err = openai.OaiStreamHandler(c, info, resp)
 	} else {
-		err, usage = openai.OpenaiHandler(c, resp, info)
+		usage, err = openai.OpenaiHandler(c, info, resp)
 	}
 	return
 }

relay/channel/dify/adaptor.go

@@ -8,6 +8,7 @@ import (
 	"one-api/dto"
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -96,11 +97,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = difyStreamHandler(c, resp, info)
+		return difyStreamHandler(c, info, resp)
 	} else {
-		err, usage = difyHandler(c, resp, info)
+		return difyHandler(c, info, resp)
 	}
 	return
 }

relay/channel/dify/relay-dify.go

@@ -14,6 +14,7 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"os"
 	"strings"
 
@@ -95,7 +96,7 @@ func uploadDifyFile(c *gin.Context, info *relaycommon.RelayInfo, user string, me
 		req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", info.ApiKey))
 
 		// Send request
-		client := service.GetImpatientHttpClient()
+		client := service.GetHttpClient()
 		resp, err := client.Do(req)
 		if err != nil {
 			common.SysError("failed to send request: " + err.Error())
@@ -209,7 +210,7 @@ func streamResponseDify2OpenAI(difyResponse DifyChunkChatCompletionResponse) *dt
 	return &response
 }
 
-func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func difyStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	var responseText string
 	usage := &dto.Usage{}
 	var nodeToken int
@@ -247,23 +248,20 @@ func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 		usage = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	}
 	usage.CompletionTokens += nodeToken
-	return nil, usage
+	return usage, nil
 }
 
-func difyHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func difyHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	var difyResponse DifyChatCompletionResponse
 	responseBody, err := io.ReadAll(resp.Body)
 
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &difyResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	fullTextResponse := dto.OpenAITextResponse{
 		Id:      difyResponse.ConversationId,
@@ -282,10 +280,10 @@ func difyHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInf
 	fullTextResponse.Choices = append(fullTextResponse.Choices, choice)
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	return nil, &difyResponse.MetaData.Usage
+	c.Writer.Write(jsonResponse)
+	return &difyResponse.MetaData.Usage, nil
 }

relay/channel/gemini/adaptor.go

@@ -11,8 +11,8 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
-	"one-api/service"
 	"one-api/setting/model_setting"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -168,30 +168,30 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.RelayMode == constant.RelayModeGemini {
 		if info.IsStream {
-			return GeminiTextGenerationStreamHandler(c, resp, info)
+			return GeminiTextGenerationStreamHandler(c, info, resp)
 		} else {
-			return GeminiTextGenerationHandler(c, resp, info)
+			return GeminiTextGenerationHandler(c, info, resp)
 		}
 	}
 
 	if strings.HasPrefix(info.UpstreamModelName, "imagen") {
-		return GeminiImageHandler(c, resp, info)
+		return GeminiImageHandler(c, info, resp)
 	}
 
 	// check if the model is an embedding model
 	if strings.HasPrefix(info.UpstreamModelName, "text-embedding") ||
 		strings.HasPrefix(info.UpstreamModelName, "embedding") ||
 		strings.HasPrefix(info.UpstreamModelName, "gemini-embedding") {
-		return GeminiEmbeddingHandler(c, resp, info)
+		return GeminiEmbeddingHandler(c, info, resp)
 	}
 
 	if info.IsStream {
-		err, usage = GeminiChatStreamHandler(c, resp, info)
+		return GeminiChatStreamHandler(c, info, resp)
 	} else {
-		err, usage = GeminiChatHandler(c, resp, info)
+		return GeminiChatHandler(c, info, resp)
 	}
 
 	//if usage.(*dto.Usage).CompletionTokenDetails.ReasoningTokens > 100 {
@@ -205,23 +205,23 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 	//	}
 	//}
 
-	return
+	return nil, types.NewError(errors.New("not implemented"), types.ErrorCodeBadResponseBody)
 }
 
-func GeminiImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func GeminiImageHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	responseBody, readErr := io.ReadAll(resp.Body)
 	if readErr != nil {
-		return nil, service.OpenAIErrorWrapper(readErr, "read_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(readErr, types.ErrorCodeBadResponseBody)
 	}
 	_ = resp.Body.Close()
 
 	var geminiResponse GeminiImageResponse
 	if jsonErr := json.Unmarshal(responseBody, &geminiResponse); jsonErr != nil {
-		return nil, service.OpenAIErrorWrapper(jsonErr, "unmarshal_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(jsonErr, types.ErrorCodeBadResponseBody)
 	}
 
 	if len(geminiResponse.Predictions) == 0 {
-		return nil, service.OpenAIErrorWrapper(errors.New("no images generated"), "no_images", http.StatusBadRequest)
+		return nil, types.NewError(errors.New("no images generated"), types.ErrorCodeBadResponseBody)
 	}
 
 	// convert to openai format response
@@ -241,7 +241,7 @@ func GeminiImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.R
 
 	jsonResponse, jsonErr := json.Marshal(openAIResponse)
 	if jsonErr != nil {
-		return nil, service.OpenAIErrorWrapper(jsonErr, "marshal_response_failed", http.StatusInternalServerError)
+		return nil, types.NewError(jsonErr, types.ErrorCodeBadResponseBody)
 	}
 
 	c.Writer.Header().Set("Content-Type", "application/json")
@@ -253,7 +253,7 @@ func GeminiImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.R
 	const imageTokens = 258
 	generatedImages := len(openAIResponse.Data)
 
-	usage = &dto.Usage{
+	usage := &dto.Usage{
 		PromptTokens:     imageTokens * generatedImages, // each generated image has fixed 258 tokens
 		CompletionTokens: 0,                             // image generation does not calculate completion tokens
 		TotalTokens:      imageTokens * generatedImages,

relay/channel/gemini/relay-gemini-native.go

@@ -1,7 +1,6 @@
 package gemini
 
 import (
-	"encoding/json"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -9,20 +8,19 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
 )
 
-func GeminiTextGenerationHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.Usage, *dto.OpenAIErrorWithStatusCode) {
+func GeminiTextGenerationHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	// 读取响应体
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return nil, service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError)
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return nil, service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 
 	if common.DebugEnabled {
@@ -31,9 +29,9 @@ func GeminiTextGenerationHandler(c *gin.Context, resp *http.Response, info *rela
 
 	// 解析为 Gemini 原生响应格式
 	var geminiResponse GeminiChatResponse
-	err = common.DecodeJson(responseBody, &geminiResponse)
+	err = common.Unmarshal(responseBody, &geminiResponse)
 	if err != nil {
-		return nil, service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 
 	// 计算使用量（基于 UsageMetadata）
@@ -54,23 +52,17 @@ func GeminiTextGenerationHandler(c *gin.Context, resp *http.Response, info *rela
 	}
 
 	// 直接返回 Gemini 原生格式的 JSON 响应
-	jsonResponse, err := json.Marshal(geminiResponse)
+	jsonResponse, err := common.Marshal(geminiResponse)
 	if err != nil {
-		return nil, service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 
-	// 设置响应头并写入响应
-	c.Writer.Header().Set("Content-Type", "application/json")
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	if err != nil {
-		return nil, service.OpenAIErrorWrapper(err, "write_response_failed", http.StatusInternalServerError)
-	}
+	common.IOCopyBytesGracefully(c, resp, jsonResponse)
 
 	return &usage, nil
 }
 
-func GeminiTextGenerationStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.Usage, *dto.OpenAIErrorWithStatusCode) {
+func GeminiTextGenerationStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	var usage = &dto.Usage{}
 	var imageCount int
 
@@ -80,7 +72,7 @@ func GeminiTextGenerationStreamHandler(c *gin.Context, resp *http.Response, info
 
 	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
 		var geminiResponse GeminiChatResponse
-		err := common.DecodeJsonStr(data, &geminiResponse)
+		err := common.UnmarshalJsonStr(data, &geminiResponse)
 		if err != nil {
 			common.LogError(c, "error unmarshalling stream response: "+err.Error())
 			return false

relay/channel/gemini/relay-gemini.go

@@ -2,6 +2,7 @@ package gemini
 
 import (
 	"encoding/json"
+	"errors"
 	"fmt"
 	"io"
 	"net/http"
@@ -12,6 +13,7 @@ import (
 	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting/model_setting"
+	"one-api/types"
 	"strconv"
 	"strings"
 	"unicode/utf8"
@@ -792,7 +794,7 @@ func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) (*dto.C
 	return &response, isStop, hasImage
 }
 
-func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func GeminiChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	// responseText := ""
 	id := helper.GetResponseID(c)
 	createAt := common.GetTimestamp()
@@ -801,7 +803,7 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 
 	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
 		var geminiResponse GeminiChatResponse
-		err := common.DecodeJsonStr(data, &geminiResponse)
+		err := common.UnmarshalJsonStr(data, &geminiResponse)
 		if err != nil {
 			common.LogError(c, "error unmarshalling stream response: "+err.Error())
 			return false
@@ -858,36 +860,25 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 	}
 	helper.Done(c)
 	//resp.Body.Close()
-	return nil, usage
+	return usage, nil
 }
 
-func GeminiChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func GeminiChatHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
+	common.CloseResponseBodyGracefully(resp)
 	if common.DebugEnabled {
 		println(string(responseBody))
 	}
 	var geminiResponse GeminiChatResponse
-	err = common.DecodeJson(responseBody, &geminiResponse)
+	err = common.Unmarshal(responseBody, &geminiResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if len(geminiResponse.Candidates) == 0 {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: "No candidates returned",
-				Type:    "server_error",
-				Param:   "",
-				Code:    500,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return nil, types.NewError(errors.New("no candidates returned"), types.ErrorCodeBadResponseBody)
 	}
 	fullTextResponse := responseGeminiChat2OpenAI(c, &geminiResponse)
 	fullTextResponse.Model = info.UpstreamModelName
@@ -911,24 +902,25 @@ func GeminiChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 	fullTextResponse.Usage = usage
 	jsonResponse, err := json.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	return nil, &usage
+	c.Writer.Write(jsonResponse)
+	return &usage, nil
 }
 
-func GeminiEmbeddingHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func GeminiEmbeddingHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	responseBody, readErr := io.ReadAll(resp.Body)
 	if readErr != nil {
-		return nil, service.OpenAIErrorWrapper(readErr, "read_response_body_failed", http.StatusInternalServerError)
+		return nil, types.NewError(readErr, types.ErrorCodeBadResponseBody)
 	}
-	_ = resp.Body.Close()
 
 	var geminiResponse GeminiEmbeddingResponse
-	if jsonErr := json.Unmarshal(responseBody, &geminiResponse); jsonErr != nil {
-		return nil, service.OpenAIErrorWrapper(jsonErr, "unmarshal_response_body_failed", http.StatusInternalServerError)
+	if jsonErr := common.Unmarshal(responseBody, &geminiResponse); jsonErr != nil {
+		return nil, types.NewError(jsonErr, types.ErrorCodeBadResponseBody)
 	}
 
 	// convert to openai format response
@@ -949,21 +941,18 @@ func GeminiEmbeddingHandler(c *gin.Context, resp *http.Response, info *relaycomm
 	// Google has not yet clarified how embedding models will be billed
 	// refer to openai billing method to use input tokens billing
 	// https://platform.openai.com/docs/guides/embeddings#what-are-embeddings
-	usage = &dto.Usage{
+	usage := &dto.Usage{
 		PromptTokens:     info.PromptTokens,
 		CompletionTokens: 0,
 		TotalTokens:      info.PromptTokens,
 	}
-	openAIResponse.Usage = *usage.(*dto.Usage)
+	openAIResponse.Usage = *usage
 
-	jsonResponse, jsonErr := json.Marshal(openAIResponse)
+	jsonResponse, jsonErr := common.Marshal(openAIResponse)
 	if jsonErr != nil {
-		return nil, service.OpenAIErrorWrapper(jsonErr, "marshal_response_failed", http.StatusInternalServerError)
+		return nil, types.NewError(jsonErr, types.ErrorCodeBadResponseBody)
 	}
 
-	c.Writer.Header().Set("Content-Type", "application/json")
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, _ = c.Writer.Write(jsonResponse)
-
+	common.IOCopyBytesGracefully(c, resp, jsonResponse)
 	return usage, nil
 }

relay/channel/jina/adaptor.go

@@ -11,6 +11,7 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/common_handler"
 	"one-api/relay/constant"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -73,11 +74,11 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 	return request, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.RelayMode == constant.RelayModeRerank {
-		err, usage = common_handler.RerankHandler(c, info, resp)
+		usage, err = common_handler.RerankHandler(c, info, resp)
 	} else if info.RelayMode == constant.RelayModeEmbeddings {
-		err, usage = openai.OpenaiHandler(c, resp, info)
+		usage, err = openai.OpenaiHandler(c, info, resp)
 	}
 	return
 }

relay/channel/jina/constant.go

@@ -3,6 +3,7 @@ package jina
 var ModelList = []string{
 	"jina-clip-v1",
 	"jina-reranker-v2-base-multilingual",
+	"jina-reranker-m0",
 }
 
 var ChannelName = "jina"

relay/channel/mistral/adaptor.go

@@ -8,6 +8,7 @@ import (
 	"one-api/relay/channel"
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -69,11 +70,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = openai.OaiStreamHandler(c, resp, info)
+		usage, err = openai.OaiStreamHandler(c, info, resp)
 	} else {
-		err, usage = openai.OpenaiHandler(c, resp, info)
+		usage, err = openai.OpenaiHandler(c, info, resp)
 	}
 	return
 }

relay/channel/mokaai/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -84,11 +85,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 
 	switch info.RelayMode {
 	case constant.RelayModeEmbeddings:
-		err, usage = mokaEmbeddingHandler(c, resp)
+		return mokaEmbeddingHandler(c, info, resp)
 	default:
 		// err, usage = mokaHandler(c, resp)
 

relay/channel/mokaai/relay-mokaai.go

@@ -2,11 +2,14 @@ package mokaai
 
 import (
 	"encoding/json"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
-	"one-api/service"
+	relaycommon "one-api/relay/common"
+	"one-api/types"
+
+	"github.com/gin-gonic/gin"
 )
 
 func embeddingRequestOpenAI2Moka(request dto.GeneralOpenAIRequest) *dto.EmbeddingRequest {
@@ -26,7 +29,7 @@ func embeddingRequestOpenAI2Moka(request dto.GeneralOpenAIRequest) *dto.Embeddin
 	}
 	return &dto.EmbeddingRequest{
 		Input: input,
-		Model:  request.Model,
+		Model: request.Model,
 	}
 }
 
@@ -47,19 +50,16 @@ func embeddingResponseMoka2OpenAI(response *dto.EmbeddingResponse) *dto.OpenAIEm
 	return &openAIEmbeddingResponse
 }
 
-func mokaEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func mokaEmbeddingHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	var baiduResponse dto.EmbeddingResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	// if baiduResponse.ErrorMsg != "" {
 	// 	return &dto.OpenAIErrorWithStatusCode{
@@ -71,13 +71,12 @@ func mokaEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 	// 	}, nil
 	// }
 	fullTextResponse := embeddingResponseMoka2OpenAI(&baiduResponse)
-	jsonResponse, err := json.Marshal(fullTextResponse)
+	jsonResponse, err := common.Marshal(fullTextResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = c.Writer.Write(jsonResponse)
-	return nil, &fullTextResponse.Usage
+	common.IOCopyBytesGracefully(c, resp, jsonResponse)
+	return &fullTextResponse.Usage, nil
 }
-

relay/channel/ollama/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -74,14 +75,14 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
-		err, usage = openai.OaiStreamHandler(c, resp, info)
+		usage, err = openai.OaiStreamHandler(c, info, resp)
 	} else {
 		if info.RelayMode == relayconstant.RelayModeEmbeddings {
-			err, usage = ollamaEmbeddingHandler(c, resp, info.PromptTokens, info.UpstreamModelName, info.RelayMode)
+			usage, err = ollamaEmbeddingHandler(c, info, resp)
 		} else {
-			err, usage = openai.OpenaiHandler(c, resp, info)
+			usage, err = openai.OpenaiHandler(c, info, resp)
 		}
 	}
 	return

relay/channel/ollama/relay-ollama.go

@@ -1,15 +1,17 @@
 package ollama
 
 import (
-	"bytes"
-	"encoding/json"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
+	relaycommon "one-api/relay/common"
 	"one-api/service"
+	"one-api/types"
 	"strings"
+
+	"github.com/gin-gonic/gin"
 )
 
 func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) (*OllamaRequest, error) {
@@ -82,22 +84,19 @@ func requestOpenAI2Embeddings(request dto.EmbeddingRequest) *OllamaEmbeddingRequ
 	}
 }
 
-func ollamaEmbeddingHandler(c *gin.Context, resp *http.Response, promptTokens int, model string, relayMode int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func ollamaEmbeddingHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	var ollamaEmbeddingResponse OllamaEmbeddingResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
-	err = resp.Body.Close()
+	common.CloseResponseBodyGracefully(resp)
+	err = common.Unmarshal(responseBody, &ollamaEmbeddingResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = json.Unmarshal(responseBody, &ollamaEmbeddingResponse)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if ollamaEmbeddingResponse.Error != "" {
-		return service.OpenAIErrorWrapper(err, "ollama_error", resp.StatusCode), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	flattenedEmbeddings := flattenEmbeddings(ollamaEmbeddingResponse.Embedding)
 	data := make([]dto.OpenAIEmbeddingResponseItem, 0, 1)
@@ -106,46 +105,22 @@ func ollamaEmbeddingHandler(c *gin.Context, resp *http.Response, promptTokens in
 		Object:    "embedding",
 	})
 	usage := &dto.Usage{
-		TotalTokens:      promptTokens,
+		TotalTokens:      info.PromptTokens,
 		CompletionTokens: 0,
-		PromptTokens:     promptTokens,
+		PromptTokens:     info.PromptTokens,
 	}
 	embeddingResponse := &dto.OpenAIEmbeddingResponse{
 		Object: "list",
 		Data:   data,
-		Model:  model,
+		Model:  info.UpstreamModelName,
 		Usage:  *usage,
 	}
-	doResponseBody, err := json.Marshal(embeddingResponse)
+	doResponseBody, err := common.Marshal(embeddingResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
-	resp.Body = io.NopCloser(bytes.NewBuffer(doResponseBody))
-	// We shouldn't set the header before we parse the response body, because the parse part may fail.
-	// And then we will have to send an error response, but in this case, the header has already been set.
-	// So the httpClient will be confused by the response.
-	// For example, Postman will report error, and we cannot check the response at all.
-	// Copy headers
-	for k, v := range resp.Header {
-		// 删除任何现有的相同头部，以防止重复添加头部
-		c.Writer.Header().Del(k)
-		for _, vv := range v {
-			c.Writer.Header().Add(k, vv)
-		}
-	}
-	// reset content length
-	c.Writer.Header().Del("Content-Length")
-	c.Writer.Header().Set("Content-Length", fmt.Sprintf("%d", len(doResponseBody)))
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = io.Copy(c.Writer, resp.Body)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	return nil, usage
+	common.IOCopyBytesGracefully(c, resp, doResponseBody)
+	return usage, nil
 }
 
 func flattenEmbeddings(embeddings [][]float64) []float64 {

relay/channel/openai/adaptor.go

@@ -9,8 +9,7 @@ import (
 	"mime/multipart"
 	"net/http"
 	"net/textproto"
-	"one-api/common"
-	constant2 "one-api/constant"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel"
 	"one-api/relay/channel/ai360"
@@ -21,8 +20,9 @@ import (
 	"one-api/relay/channel/xinference"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/common_handler"
-	"one-api/relay/constant"
+	relayconstant "one-api/relay/constant"
 	"one-api/service"
+	"one-api/types"
 	"path/filepath"
 	"strings"
 
@@ -54,7 +54,7 @@ func (a *Adaptor) Init(info *relaycommon.RelayInfo) {
 	a.ChannelType = info.ChannelType
 
 	// initialize ThinkingContentInfo when thinking_to_content is enabled
-	if think2Content, ok := info.ChannelSetting[constant2.ChannelSettingThinkingToContent].(bool); ok && think2Content {
+	if info.ChannelSetting.ThinkingToContent {
 		info.ThinkingContentInfo = relaycommon.ThinkingContentInfo{
 			IsFirstThinkingContent:  true,
 			SendLastThinkingContent: false,
@@ -67,7 +67,7 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 	if info.RelayFormat == relaycommon.RelayFormatClaude {
 		return fmt.Sprintf("%s/v1/chat/completions", info.BaseUrl), nil
 	}
-	if info.RelayMode == constant.RelayModeRealtime {
+	if info.RelayMode == relayconstant.RelayModeRealtime {
 		if strings.HasPrefix(info.BaseUrl, "https://") {
 			baseUrl := strings.TrimPrefix(info.BaseUrl, "https://")
 			baseUrl = "wss://" + baseUrl
@@ -79,10 +79,10 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		}
 	}
 	switch info.ChannelType {
-	case common.ChannelTypeAzure:
+	case constant.ChannelTypeAzure:
 		apiVersion := info.ApiVersion
 		if apiVersion == "" {
-			apiVersion = constant2.AzureDefaultAPIVersion
+			apiVersion = constant.AzureDefaultAPIVersion
 		}
 		// https://learn.microsoft.com/en-us/azure/cognitive-services/openai/chatgpt-quickstart?pivots=rest-api&tabs=command-line#rest-api
 		requestURL := strings.Split(info.RequestURLPath, "?")[0]
@@ -90,25 +90,25 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		task := strings.TrimPrefix(requestURL, "/v1/")
 
 		// 特殊处理 responses API
-		if info.RelayMode == constant.RelayModeResponses {
+		if info.RelayMode == relayconstant.RelayModeResponses {
 			requestURL = fmt.Sprintf("/openai/v1/responses?api-version=preview")
 			return relaycommon.GetFullRequestURL(info.BaseUrl, requestURL, info.ChannelType), nil
 		}
 
 		model_ := info.UpstreamModelName
 		// 2025年5月10日后创建的渠道不移除.
-		if info.ChannelCreateTime < constant2.AzureNoRemoveDotTime {
+		if info.ChannelCreateTime < constant.AzureNoRemoveDotTime {
 			model_ = strings.Replace(model_, ".", "", -1)
 		}
 		// https://github.com/songquanpeng/one-api/issues/67
 		requestURL = fmt.Sprintf("/openai/deployments/%s/%s", model_, task)
-		if info.RelayMode == constant.RelayModeRealtime {
+		if info.RelayMode == relayconstant.RelayModeRealtime {
 			requestURL = fmt.Sprintf("/openai/realtime?deployment=%s&api-version=%s", model_, apiVersion)
 		}
 		return relaycommon.GetFullRequestURL(info.BaseUrl, requestURL, info.ChannelType), nil
-	case common.ChannelTypeMiniMax:
+	case constant.ChannelTypeMiniMax:
 		return minimax.GetRequestURL(info)
-	case common.ChannelTypeCustom:
+	case constant.ChannelTypeCustom:
 		url := info.BaseUrl
 		url = strings.Replace(url, "{model}", info.UpstreamModelName, -1)
 		return url, nil
@@ -119,14 +119,14 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, header *http.Header, info *relaycommon.RelayInfo) error {
 	channel.SetupApiRequestHeader(info, c, header)
-	if info.ChannelType == common.ChannelTypeAzure {
+	if info.ChannelType == constant.ChannelTypeAzure {
 		header.Set("api-key", info.ApiKey)
 		return nil
 	}
-	if info.ChannelType == common.ChannelTypeOpenAI && "" != info.Organization {
+	if info.ChannelType == constant.ChannelTypeOpenAI && "" != info.Organization {
 		header.Set("OpenAI-Organization", info.Organization)
 	}
-	if info.RelayMode == constant.RelayModeRealtime {
+	if info.RelayMode == relayconstant.RelayModeRealtime {
 		swp := c.Request.Header.Get("Sec-WebSocket-Protocol")
 		if swp != "" {
 			items := []string{
@@ -145,8 +145,8 @@ func (a *Adaptor) SetupRequestHeader(c *gin.Context, header *http.Header, info *
 	} else {
 		header.Set("Authorization", "Bearer "+info.ApiKey)
 	}
-	if info.ChannelType == common.ChannelTypeOpenRouter {
-		header.Set("HTTP-Referer", "https://github.com/Calcium-Ion/new-api")
+	if info.ChannelType == constant.ChannelTypeOpenRouter {
+		header.Set("HTTP-Referer", "https://www.newapi.ai")
 		header.Set("X-Title", "New API")
 	}
 	return nil
@@ -156,10 +156,10 @@ func (a *Adaptor) ConvertOpenAIRequest(c *gin.Context, info *relaycommon.RelayIn
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	if info.ChannelType != common.ChannelTypeOpenAI && info.ChannelType != common.ChannelTypeAzure {
+	if info.ChannelType != constant.ChannelTypeOpenAI && info.ChannelType != constant.ChannelTypeAzure {
 		request.StreamOptions = nil
 	}
-	if info.ChannelType == common.ChannelTypeOpenRouter {
+	if info.ChannelType == constant.ChannelTypeOpenRouter {
 		if len(request.Usage) == 0 {
 			request.Usage = json.RawMessage(`{"include":true}`)
 		}
@@ -205,7 +205,7 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 
 func (a *Adaptor) ConvertAudioRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.AudioRequest) (io.Reader, error) {
 	a.ResponseFormat = request.ResponseFormat
-	if info.RelayMode == constant.RelayModeAudioSpeech {
+	if info.RelayMode == relayconstant.RelayModeAudioSpeech {
 		jsonData, err := json.Marshal(request)
 		if err != nil {
 			return nil, fmt.Errorf("error marshalling object: %w", err)
@@ -254,7 +254,7 @@ func (a *Adaptor) ConvertAudioRequest(c *gin.Context, info *relaycommon.RelayInf
 
 func (a *Adaptor) ConvertImageRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.ImageRequest) (any, error) {
 	switch info.RelayMode {
-	case constant.RelayModeImagesEdits:
+	case relayconstant.RelayModeImagesEdits:
 
 		var requestBody bytes.Buffer
 		writer := multipart.NewWriter(&requestBody)
@@ -411,42 +411,42 @@ func (a *Adaptor) ConvertOpenAIResponsesRequest(c *gin.Context, info *relaycommo
 }
 
 func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (any, error) {
-	if info.RelayMode == constant.RelayModeAudioTranscription ||
-		info.RelayMode == constant.RelayModeAudioTranslation ||
-		info.RelayMode == constant.RelayModeImagesEdits {
+	if info.RelayMode == relayconstant.RelayModeAudioTranscription ||
+		info.RelayMode == relayconstant.RelayModeAudioTranslation ||
+		info.RelayMode == relayconstant.RelayModeImagesEdits {
 		return channel.DoFormRequest(a, c, info, requestBody)
-	} else if info.RelayMode == constant.RelayModeRealtime {
+	} else if info.RelayMode == relayconstant.RelayModeRealtime {
 		return channel.DoWssRequest(a, c, info, requestBody)
 	} else {
 		return channel.DoApiRequest(a, c, info, requestBody)
 	}
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayMode {
-	case constant.RelayModeRealtime:
+	case relayconstant.RelayModeRealtime:
 		err, usage = OpenaiRealtimeHandler(c, info)
-	case constant.RelayModeAudioSpeech:
-		err, usage = OpenaiTTSHandler(c, resp, info)
-	case constant.RelayModeAudioTranslation:
+	case relayconstant.RelayModeAudioSpeech:
+		usage = OpenaiTTSHandler(c, resp, info)
+	case relayconstant.RelayModeAudioTranslation:
 		fallthrough
-	case constant.RelayModeAudioTranscription:
+	case relayconstant.RelayModeAudioTranscription:
 		err, usage = OpenaiSTTHandler(c, resp, info, a.ResponseFormat)
-	case constant.RelayModeImagesGenerations, constant.RelayModeImagesEdits:
-		err, usage = OpenaiHandlerWithUsage(c, resp, info)
-	case constant.RelayModeRerank:
-		err, usage = common_handler.RerankHandler(c, info, resp)
-	case constant.RelayModeResponses:
+	case relayconstant.RelayModeImagesGenerations, relayconstant.RelayModeImagesEdits:
+		usage, err = OpenaiHandlerWithUsage(c, info, resp)
+	case relayconstant.RelayModeRerank:
+		usage, err = common_handler.RerankHandler(c, info, resp)
+	case relayconstant.RelayModeResponses:
 		if info.IsStream {
-			err, usage = OaiResponsesStreamHandler(c, resp, info)
+			usage, err = OaiResponsesStreamHandler(c, info, resp)
 		} else {
-			err, usage = OaiResponsesHandler(c, resp, info)
+			usage, err = OaiResponsesHandler(c, info, resp)
 		}
 	default:
 		if info.IsStream {
-			err, usage = OaiStreamHandler(c, resp, info)
+			usage, err = OaiStreamHandler(c, info, resp)
 		} else {
-			err, usage = OpenaiHandler(c, resp, info)
+			usage, err = OpenaiHandler(c, info, resp)
 		}
 	}
 	return
@@ -454,17 +454,17 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 
 func (a *Adaptor) GetModelList() []string {
 	switch a.ChannelType {
-	case common.ChannelType360:
+	case constant.ChannelType360:
 		return ai360.ModelList
-	case common.ChannelTypeMoonshot:
+	case constant.ChannelTypeMoonshot:
 		return moonshot.ModelList
-	case common.ChannelTypeLingYiWanWu:
+	case constant.ChannelTypeLingYiWanWu:
 		return lingyiwanwu.ModelList
-	case common.ChannelTypeMiniMax:
+	case constant.ChannelTypeMiniMax:
 		return minimax.ModelList
-	case common.ChannelTypeXinference:
+	case constant.ChannelTypeXinference:
 		return xinference.ModelList
-	case common.ChannelTypeOpenRouter:
+	case constant.ChannelTypeOpenRouter:
 		return openrouter.ModelList
 	default:
 		return ModelList
@@ -473,17 +473,17 @@ func (a *Adaptor) GetModelList() []string {
 
 func (a *Adaptor) GetChannelName() string {
 	switch a.ChannelType {
-	case common.ChannelType360:
+	case constant.ChannelType360:
 		return ai360.ChannelName
-	case common.ChannelTypeMoonshot:
+	case constant.ChannelTypeMoonshot:
 		return moonshot.ChannelName
-	case common.ChannelTypeLingYiWanWu:
+	case constant.ChannelTypeLingYiWanWu:
 		return lingyiwanwu.ChannelName
-	case common.ChannelTypeMiniMax:
+	case constant.ChannelTypeMiniMax:
 		return minimax.ChannelName
-	case common.ChannelTypeXinference:
+	case constant.ChannelTypeXinference:
 		return xinference.ChannelName
-	case common.ChannelTypeOpenRouter:
+	case constant.ChannelTypeOpenRouter:
 		return openrouter.ChannelName
 	default:
 		return ChannelName

relay/channel/openai/relay-openai.go

@@ -2,7 +2,6 @@ package openai
 
 import (
 	"bytes"
-	"encoding/json"
 	"fmt"
 	"io"
 	"math"
@@ -18,6 +17,8 @@ import (
 	"path/filepath"
 	"strings"
 
+	"one-api/types"
+
 	"github.com/bytedance/gopkg/util/gopool"
 	"github.com/gin-gonic/gin"
 	"github.com/gorilla/websocket"
@@ -34,7 +35,7 @@ func sendStreamData(c *gin.Context, info *relaycommon.RelayInfo, data string, fo
 	}
 
 	var lastStreamResponse dto.ChatCompletionsStreamResponse
-	if err := common.DecodeJsonStr(data, &lastStreamResponse); err != nil {
+	if err := common.UnmarshalJsonStr(data, &lastStreamResponse); err != nil {
 		return err
 	}
 
@@ -105,18 +106,19 @@ func sendStreamData(c *gin.Context, info *relaycommon.RelayInfo, data string, fo
 	return helper.ObjectData(c, lastStreamResponse)
 }
 
-func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OaiStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	if resp == nil || resp.Body == nil {
 		common.LogError(c, "invalid response or response body")
-		return service.OpenAIErrorWrapper(fmt.Errorf("invalid response"), "invalid_response", http.StatusInternalServerError), nil
+		return nil, types.NewError(fmt.Errorf("invalid response"), types.ErrorCodeBadResponse)
 	}
 
-	containStreamUsage := false
+	defer common.CloseResponseBodyGracefully(resp)
+
+	model := info.UpstreamModelName
 	var responseId string
 	var createAt int64 = 0
 	var systemFingerprint string
-	model := info.UpstreamModelName
-
+	var containStreamUsage bool
 	var responseTextBuilder strings.Builder
 	var toolCount int
 	var usage = &dto.Usage{}
@@ -124,12 +126,12 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	var forceFormat bool
 	var thinkToContent bool
 
-	if forceFmt, ok := info.ChannelSetting[constant.ForceFormat].(bool); ok {
-		forceFormat = forceFmt
+	if info.ChannelSetting.ForceFormat {
+		forceFormat = true
 	}
 
-	if think2Content, ok := info.ChannelSetting[constant.ChannelSettingThinkingToContent].(bool); ok {
-		thinkToContent = think2Content
+	if info.ChannelSetting.ThinkingToContent {
+		thinkToContent = true
 	}
 
 	var (
@@ -148,31 +150,15 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 		return true
 	})
 
+	// 处理最后的响应
 	shouldSendLastResp := true
-	var lastStreamResponse dto.ChatCompletionsStreamResponse
-	err := common.DecodeJsonStr(lastStreamData, &lastStreamResponse)
-	if err == nil {
-		responseId = lastStreamResponse.Id
-		createAt = lastStreamResponse.Created
-		systemFingerprint = lastStreamResponse.GetSystemFingerprint()
-		model = lastStreamResponse.Model
-		if service.ValidUsage(lastStreamResponse.Usage) {
-			containStreamUsage = true
-			usage = lastStreamResponse.Usage
-			if !info.ShouldIncludeUsage {
-				shouldSendLastResp = false
-			}
-		}
-		for _, choice := range lastStreamResponse.Choices {
-			if choice.FinishReason != nil {
-				shouldSendLastResp = true
-			}
-		}
+	if err := handleLastResponse(lastStreamData, &responseId, &createAt, &systemFingerprint, &model, &usage,
+		&containStreamUsage, info, &shouldSendLastResp); err != nil {
+		common.SysError("error handling last response: " + err.Error())
 	}
 
-	if shouldSendLastResp {
-		sendStreamData(c, info, lastStreamData, forceFormat, thinkToContent)
-		//err = handleStreamFormat(c, info, lastStreamData, forceFormat, thinkToContent)
+	if shouldSendLastResp && info.RelayFormat == relaycommon.RelayFormatOpenAI {
+		_ = sendStreamData(c, info, lastStreamData, forceFormat, thinkToContent)
 	}
 
 	// 处理token计算
@@ -184,7 +170,7 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 		usage = service.ResponseText2Usage(responseTextBuilder.String(), info.UpstreamModelName, info.PromptTokens)
 		usage.CompletionTokens += toolCount * 7
 	} else {
-		if info.ChannelType == common.ChannelTypeDeepSeek {
+		if info.ChannelType == constant.ChannelTypeDeepSeek {
 			if usage.PromptCacheHitTokens != 0 {
 				usage.PromptTokensDetails.CachedTokens = usage.PromptCacheHitTokens
 			}
@@ -193,33 +179,28 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 
 	handleFinalResponse(c, info, lastStreamData, responseId, createAt, model, systemFingerprint, usage, containStreamUsage)
 
-	return nil, usage
+	return usage, nil
 }
 
-func OpenaiHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OpenaiHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	var simpleResponse dto.OpenAITextResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeReadResponseBodyFailed)
 	}
-	err = resp.Body.Close()
+	err = common.Unmarshal(responseBody, &simpleResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = common.DecodeJson(responseBody, &simpleResponse)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if simpleResponse.Error != nil && simpleResponse.Error.Type != "" {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error:      *simpleResponse.Error,
-			StatusCode: resp.StatusCode,
-		}, nil
+		return nil, types.WithOpenAIError(*simpleResponse.Error, resp.StatusCode)
 	}
 
 	forceFormat := false
-	if forceFmt, ok := info.ChannelSetting[constant.ForceFormat].(bool); ok {
-		forceFormat = forceFmt
+	if info.ChannelSetting.ForceFormat {
+		forceFormat = true
 	}
 
 	if simpleResponse.Usage.TotalTokens == 0 || (simpleResponse.Usage.PromptTokens == 0 && simpleResponse.Usage.CompletionTokens == 0) {
@@ -238,49 +219,35 @@ func OpenaiHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayI
 	switch info.RelayFormat {
 	case relaycommon.RelayFormatOpenAI:
 		if forceFormat {
-			responseBody, err = json.Marshal(simpleResponse)
+			responseBody, err = common.Marshal(simpleResponse)
 			if err != nil {
-				return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+				return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 			}
 		} else {
 			break
 		}
 	case relaycommon.RelayFormatClaude:
 		claudeResp := service.ResponseOpenAI2Claude(&simpleResponse, info)
-		claudeRespStr, err := json.Marshal(claudeResp)
+		claudeRespStr, err := common.Marshal(claudeResp)
 		if err != nil {
-			return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+			return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 		}
 		responseBody = claudeRespStr
 	}
 
-	// Reset response body
-	resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
-	// We shouldn't set the header before we parse the response body, because the parse part may fail.
-	// And then we will have to send an error response, but in this case, the header has already been set.
-	// So the httpClient will be confused by the response.
-	// For example, Postman will report error, and we cannot check the response at all.
-	for k, v := range resp.Header {
-		c.Writer.Header().Set(k, v[0])
-	}
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = io.Copy(c.Writer, resp.Body)
-	if err != nil {
-		//return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
-		common.SysError("error copying response body: " + err.Error())
-	}
-	resp.Body.Close()
-	return nil, &simpleResponse.Usage
+	common.IOCopyBytesGracefully(c, resp, responseBody)
+
+	return &simpleResponse.Usage, nil
 }
 
-func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) *dto.Usage {
 	// the status code has been judged before, if there is a body reading failure,
 	// it should be regarded as a non-recoverable error, so it should not return err for external retry.
 	// Analogous to nginx's load balancing, it will only retry if it can't be requested or
 	// if the upstream returns a specific status code, once the upstream has already written the header,
 	// the subsequent failure of the response body should be regarded as a non-recoverable error,
 	// and can be terminated directly.
-	defer resp.Body.Close()
+	defer common.CloseResponseBodyGracefully(resp)
 	usage := &dto.Usage{}
 	usage.PromptTokens = info.PromptTokens
 	usage.TotalTokens = info.PromptTokens
@@ -293,38 +260,23 @@ func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	if err != nil {
 		common.LogError(c, err.Error())
 	}
-	return nil, usage
+	return usage
 }
 
-func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, responseFormat string) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, responseFormat string) (*types.NewAPIError, *dto.Usage) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	// count tokens by audio file duration
 	audioTokens, err := countAudioTokens(c)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "count_audio_tokens_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeCountTokenFailed), nil
 	}
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+		return types.NewError(err, types.ErrorCodeReadResponseBodyFailed), nil
 	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	// Reset response body
-	resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
-	// We shouldn't set the header before we parse the response body, because the parse part may fail.
-	// And then we will have to send an error response, but in this case, the header has already been set.
-	// So the httpClient will be confused by the response.
-	// For example, Postman will report error, and we cannot check the response at all.
-	for k, v := range resp.Header {
-		c.Writer.Header().Set(k, v[0])
-	}
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = io.Copy(c.Writer, resp.Body)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
-	}
-	resp.Body.Close()
+	// 写入新的 response body
+	common.IOCopyBytesGracefully(c, resp, responseBody)
 
 	usage := &dto.Usage{}
 	usage.PromptTokens = audioTokens
@@ -375,9 +327,9 @@ func countAudioTokens(c *gin.Context) (int, error) {
 	return int(math.Round(math.Ceil(duration) / 60.0 * 1000)), nil // 1 minute 相当于 1k tokens
 }
 
-func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.RealtimeUsage) {
+func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*types.NewAPIError, *dto.RealtimeUsage) {
 	if info == nil || info.ClientWs == nil || info.TargetWs == nil {
-		return service.OpenAIErrorWrapper(fmt.Errorf("invalid websocket connection"), "invalid_connection", http.StatusBadRequest), nil
+		return types.NewError(fmt.Errorf("invalid websocket connection"), types.ErrorCodeBadResponse), nil
 	}
 
 	info.IsStream = true
@@ -415,7 +367,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 				}
 
 				realtimeEvent := &dto.RealtimeEvent{}
-				err = json.Unmarshal(message, realtimeEvent)
+				err = common.Unmarshal(message, realtimeEvent)
 				if err != nil {
 					errChan <- fmt.Errorf("error unmarshalling message: %v", err)
 					return
@@ -475,7 +427,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 				}
 				info.SetFirstResponseTime()
 				realtimeEvent := &dto.RealtimeEvent{}
-				err = json.Unmarshal(message, realtimeEvent)
+				err = common.Unmarshal(message, realtimeEvent)
 				if err != nil {
 					errChan <- fmt.Errorf("error unmarshalling message: %v", err)
 					return
@@ -522,9 +474,9 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 						localUsage = &dto.RealtimeUsage{}
 						// print now usage
 					}
-					//common.LogInfo(c, fmt.Sprintf("realtime streaming sumUsage: %v", sumUsage))
-					//common.LogInfo(c, fmt.Sprintf("realtime streaming localUsage: %v", localUsage))
-					//common.LogInfo(c, fmt.Sprintf("realtime streaming localUsage: %v", localUsage))
+					common.LogInfo(c, fmt.Sprintf("realtime streaming sumUsage: %v", sumUsage))
+					common.LogInfo(c, fmt.Sprintf("realtime streaming localUsage: %v", localUsage))
+					common.LogInfo(c, fmt.Sprintf("realtime streaming localUsage: %v", localUsage))
 
 				} else if realtimeEvent.Type == dto.RealtimeEventTypeSessionUpdated || realtimeEvent.Type == dto.RealtimeEventTypeSessionCreated {
 					realtimeSession := realtimeEvent.Session
@@ -600,41 +552,26 @@ func preConsumeUsage(ctx *gin.Context, info *relaycommon.RelayInfo, usage *dto.R
 	return err
 }
 
-func OpenaiHandlerWithUsage(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OpenaiHandlerWithUsage(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	// Reset response body
-	resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
-	// We shouldn't set the header before we parse the response body, because the parse part may fail.
-	// And then we will have to send an error response, but in this case, the header has already been set.
-	// So the httpClient will be confused by the response.
-	// For example, Postman will report error, and we cannot check the response at all.
-	for k, v := range resp.Header {
-		c.Writer.Header().Set(k, v[0])
-	}
-	// reset content length
-	c.Writer.Header().Set("Content-Length", fmt.Sprintf("%d", len(responseBody)))
-	c.Writer.WriteHeader(resp.StatusCode)
-	_, err = io.Copy(c.Writer, resp.Body)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeReadResponseBodyFailed)
 	}
 
 	var usageResp dto.SimpleResponse
-	err = json.Unmarshal(responseBody, &usageResp)
+	err = common.Unmarshal(responseBody, &usageResp)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "parse_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
+
+	// 写入新的 response body
+	common.IOCopyBytesGracefully(c, resp, responseBody)
+
+	// Once we've written to the client, we should not return errors anymore
+	// because the upstream has already consumed resources and returned content
+	// We should still perform billing even if parsing fails
 	// format
 	if usageResp.InputTokens > 0 {
 		usageResp.PromptTokens += usageResp.InputTokens
@@ -646,5 +583,5 @@ func OpenaiHandlerWithUsage(c *gin.Context, resp *http.Response, info *relaycomm
 		usageResp.PromptTokensDetails.ImageTokens += usageResp.InputTokensDetails.ImageTokens
 		usageResp.PromptTokensDetails.TextTokens += usageResp.InputTokensDetails.TextTokens
 	}
-	return nil, &usageResp.Usage
+	return &usageResp.Usage, nil
 }

relay/channel/openai/relay_responses.go

@@ -1,7 +1,6 @@
 package openai
 
 import (
-	"bytes"
 	"fmt"
 	"io"
 	"net/http"
@@ -10,53 +9,32 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
+	"one-api/types"
 	"strings"
 
 	"github.com/gin-gonic/gin"
 )
 
-func OaiResponsesHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OaiResponsesHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	// read response body
 	var responsesResponse dto.OpenAIResponsesResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeReadResponseBodyFailed)
 	}
-	err = resp.Body.Close()
+	err = common.Unmarshal(responseBody, &responsesResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = common.DecodeJson(responseBody, &responsesResponse)
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+		return nil, types.NewError(err, types.ErrorCodeBadResponseBody)
 	}
 	if responsesResponse.Error != nil {
-		return &dto.OpenAIErrorWithStatusCode{
-			Error: dto.OpenAIError{
-				Message: responsesResponse.Error.Message,
-				Type:    "openai_error",
-				Code:    responsesResponse.Error.Code,
-			},
-			StatusCode: resp.StatusCode,
-		}, nil
+		return nil, types.WithOpenAIError(*responsesResponse.Error, resp.StatusCode)
 	}
 
-	// reset response body
-	resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
-	// We shouldn't set the header before we parse the response body, because the parse part may fail.
-	// And then we will have to send an error response, but in this case, the header has already been set.
-	// So the httpClient will be confused by the response.
-	// For example, Postman will report error, and we cannot check the response at all.
-	for k, v := range resp.Header {
-		c.Writer.Header().Set(k, v[0])
-	}
-	c.Writer.WriteHeader(resp.StatusCode)
-	// copy response body
-	_, err = io.Copy(c.Writer, resp.Body)
-	if err != nil {
-		common.SysError("error copying response body: " + err.Error())
-	}
-	resp.Body.Close()
+	// 写入新的 response body
+	common.IOCopyBytesGracefully(c, resp, responseBody)
+
 	// compute usage
 	usage := dto.Usage{}
 	usage.PromptTokens = responsesResponse.Usage.InputTokens
@@ -66,13 +44,13 @@ func OaiResponsesHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 	for _, tool := range responsesResponse.Tools {
 		info.ResponsesUsageInfo.BuiltInTools[tool.Type].CallCount++
 	}
-	return nil, &usage
+	return &usage, nil
 }
 
-func OaiResponsesStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+func OaiResponsesStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
 	if resp == nil || resp.Body == nil {
 		common.LogError(c, "invalid response or response body")
-		return service.OpenAIErrorWrapper(fmt.Errorf("invalid response"), "invalid_response", http.StatusInternalServerError), nil
+		return nil, types.NewError(fmt.Errorf("invalid response"), types.ErrorCodeBadResponse)
 	}
 
 	var usage = &dto.Usage{}
@@ -82,7 +60,7 @@ func OaiResponsesStreamHandler(c *gin.Context, resp *http.Response, info *relayc
 
 		// 检查当前数据是否包含 completed 状态和 usage 信息
 		var streamResponse dto.ResponsesStreamResponse
-		if err := common.DecodeJsonStr(data, &streamResponse); err == nil {
+		if err := common.UnmarshalJsonStr(data, &streamResponse); err == nil {
 			sendResponsesStreamData(c, streamResponse, data)
 			switch streamResponse.Type {
 			case "response.completed":
@@ -115,5 +93,5 @@ func OaiResponsesStreamHandler(c *gin.Context, resp *http.Response, info *relayc
 		}
 	}
 
-	return nil, usage
+	return usage, nil
 }

relay/channel/palm/adaptor.go

@@ -9,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
+	"one-api/types"
 
 	"github.com/gin-gonic/gin"
 )
@@ -70,13 +71,13 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.IsStream {
 		var responseText string
 		err, responseText = palmStreamHandler(c, resp)
 		usage = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	} else {
-		err, usage = palmHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		usage, err = palmHandler(c, info, resp)
 	}
 	return
 }