✨ feat: add pull request template and enforce branching strategy in workflow

fix: ali embedding lose prompt_tokens

.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=90
 
 
 # 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."

Dockerfile

@@ -24,8 +24,7 @@ RUN go build -ldflags "-s -w -X 'one-api/common.Version=$(cat VERSION)'" -o one-
 
 FROM alpine
 
-RUN apk update \
-    && apk upgrade \
+RUN apk upgrade --no-cache \
     && apk add --no-cache ca-certificates tzdata ffmpeg \
     && update-ca-certificates
 

README.en.md

@@ -100,7 +100,7 @@ This version supports multiple models, please refer to [API Documentation-Relay
 For detailed configuration instructions, please refer to [Installation Guide-Environment Variables Configuration](https://docs.newapi.pro/installation/environment-variables):
 
 - `GENERATE_DEFAULT_TOKEN`: Whether to generate initial tokens for newly registered users, default is `false`
-- `STREAMING_TIMEOUT`: Streaming response timeout, default is 60 seconds
+- `STREAMING_TIMEOUT`: Streaming response timeout, default is 120 seconds
 - `DIFY_DEBUG`: Whether to output workflow and node information for Dify channels, default is `true`
 - `FORCE_STREAM_OPTION`: Whether to override client stream_options parameter, default is `true`
 - `GET_MEDIA_TOKEN`: Whether to count image tokens, default is `true`

README.md

@@ -103,7 +103,7 @@ New API提供了丰富的功能，详细特性请参考[特性说明](https://do
 详细配置说明请参考[安装指南-环境变量配置](https://docs.newapi.pro/installation/environment-variables)：
 
 - `GENERATE_DEFAULT_TOKEN`：是否为新注册用户生成初始令牌，默认为 `false`
-- `STREAMING_TIMEOUT`：流式回复超时时间，默认60秒
+- `STREAMING_TIMEOUT`：流式回复超时时间，默认120秒
 - `DIFY_DEBUG`：Dify渠道是否输出工作流和节点信息，默认 `true`
 - `FORCE_STREAM_OPTION`：是否覆盖客户端stream_options参数，默认 `true`
 - `GET_MEDIA_TOKEN`：是否统计图片token，默认 `true`

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 +1,29 @@
+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.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}
+		}
+	}
+	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 = UnmarshalJson(requestBody, &v)
 	} else {
 		// skip for now
 		// TODO: someday non json request have variant model, we will need to implementation this
@@ -43,3 +44,35 @@ 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))
+}

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

@@ -5,12 +5,16 @@ import (
 	"encoding/json"
 )
 
-func DecodeJson(data []byte, v any) error {
-	return json.NewDecoder(bytes.NewReader(data)).Decode(v)
+func UnmarshalJson(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 DecodeJson(reader *bytes.Reader, v any) error {
+	return json.NewDecoder(reader).Decode(v)
 }
 
 func EncodeJson(v any) ([]byte, error) {

common/model.go

@@ -0,0 +1,21 @@
+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",
+	}
+)
+
+func IsOpenAIResponseOnlyModel(modelName string) bool {
+	for _, m := range OpenAIResponseOnlyModels {
+		if strings.Contains(m, modelName) {
+			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") == "" {

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/context_key.go

@@ -1,10 +1,35 @@
 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"
+	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 */
+	ContextKeyBaseUrl        ContextKey = "base_url"
+	ContextKeyChannelType    ContextKey = "channel_type"
+	ContextKeyChannelId      ContextKey = "channel_id"
+	ContextKeyChannelSetting ContextKey = "channel_setting"
+	ContextKeyParamOverride  ContextKey = "param_override"
+
+	/* 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 +1,11 @@
+package constant
+
+type EndpointType string
+
+const (
+	EndpointTypeOpenAI         EndpointType = "openai"
+	EndpointTypeOpenAIResponse EndpointType = "openai-response"
+	EndpointTypeAnthropic      EndpointType = "anthropic"
+	EndpointTypeGemini         EndpointType = "gemini"
+	EndpointTypeJinaRerank     EndpointType = "jina-rerank"
+)

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", 60)
-	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/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{

controller/channel-billing.go

@@ -4,11 +4,14 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/shopspring/decimal"
 	"io"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/model"
 	"one-api/service"
+	"one-api/setting"
 	"strconv"
 	"time"
 
@@ -304,34 +307,70 @@ func updateChannelOpenRouterBalance(channel *model.Channel) (float64, error) {
 	return balance, nil
 }
 
+func updateChannelMoonshotBalance(channel *model.Channel) (float64, error) {
+	url := "https://api.moonshot.cn/v1/users/me/balance"
+	body, err := GetResponseBody("GET", url, channel, GetAuthHeader(channel.Key))
+	if err != nil {
+		return 0, err
+	}
+
+	type MoonshotBalanceData struct {
+		AvailableBalance float64 `json:"available_balance"`
+		VoucherBalance   float64 `json:"voucher_balance"`
+		CashBalance      float64 `json:"cash_balance"`
+	}
+
+	type MoonshotBalanceResponse struct {
+		Code   int                 `json:"code"`
+		Data   MoonshotBalanceData `json:"data"`
+		Scode  string              `json:"scode"`
+		Status bool                `json:"status"`
+	}
+
+	response := MoonshotBalanceResponse{}
+	err = json.Unmarshal(body, &response)
+	if err != nil {
+		return 0, err
+	}
+	if !response.Status || response.Code != 0 {
+		return 0, fmt.Errorf("failed to update moonshot balance, status: %v, code: %d, scode: %s", response.Status, response.Code, response.Scode)
+	}
+	availableBalanceCny := response.Data.AvailableBalance
+	availableBalanceUsd := decimal.NewFromFloat(availableBalanceCny).Div(decimal.NewFromFloat(setting.Price)).InexactFloat64()
+	channel.UpdateBalance(availableBalanceUsd)
+	return availableBalanceUsd, nil
+}
+
 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 constant.ChannelTypeMoonshot:
+		return updateChannelMoonshotBalance(channel)
 	default:
 		return 0, errors.New("尚未实现")
 	}

controller/channel-test.go

@@ -11,12 +11,12 @@ 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"
 	"strconv"
@@ -31,18 +31,21 @@ import (
 
 func testChannel(channel *model.Channel, testModel string) (err error, openAIErrorWithStatusCode *dto.OpenAIErrorWithStatusCode) {
 	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 +56,7 @@ 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
 	}
 
@@ -99,7 +102,7 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	}
 	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
@@ -171,7 +174,7 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	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.Group, other)
+		quota, "模型测试", 0, quota, int(consumedTime), false, info.UsingGroup, other)
 	common.SysLog(fmt.Sprintf("testing channel #%d, response: \n%s", channel.Id, string(respBody)))
 	return nil, nil
 }

controller/channel.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/model"
 	"strconv"
 	"strings"
@@ -40,6 +41,17 @@ type OpenAIModelsResponse struct {
 	Success bool          `json:"success"`
 }
 
+func parseStatusFilter(statusParam string) int {
+	switch strings.ToLower(statusParam) {
+	case "enabled", "1":
+		return common.ChannelStatusEnabled
+	case "disabled", "0":
+		return 0
+	default:
+		return -1
+	}
+}
+
 func GetAllChannels(c *gin.Context) {
 	p, _ := strconv.Atoi(c.Query("p"))
 	pageSize, _ := strconv.Atoi(c.Query("page_size"))
@@ -52,6 +64,9 @@ func GetAllChannels(c *gin.Context) {
 	channelData := make([]*model.Channel, 0)
 	idSort, _ := strconv.ParseBool(c.Query("id_sort"))
 	enableTagMode, _ := strconv.ParseBool(c.Query("tag_mode"))
+	statusParam := c.Query("status")
+	// statusFilter: -1 all, 1 enabled, 0 disabled (include auto & manual)
+	statusFilter := parseStatusFilter(statusParam)
 	// type filter
 	typeStr := c.Query("type")
 	typeFilter := -1
@@ -64,42 +79,75 @@ func GetAllChannels(c *gin.Context) {
 	var total int64
 
 	if enableTagMode {
-		// tag 分页：先分页 tag，再取各 tag 下 channels
 		tags, err := model.GetPaginatedTags((p-1)*pageSize, pageSize)
 		if err != nil {
 			c.JSON(http.StatusOK, gin.H{"success": false, "message": err.Error()})
 			return
 		}
 		for _, tag := range tags {
-			if tag != nil && *tag != "" {
-				tagChannel, err := model.GetChannelsByTag(*tag, idSort)
-				if err == nil {
-					channelData = append(channelData, tagChannel...)
-				}
+			if tag == nil || *tag == "" {
+				continue
 			}
+			tagChannels, err := model.GetChannelsByTag(*tag, idSort)
+			if err != nil {
+				continue
+			}
+			filtered := make([]*model.Channel, 0)
+			for _, ch := range tagChannels {
+				if statusFilter == common.ChannelStatusEnabled && ch.Status != common.ChannelStatusEnabled {
+					continue
+				}
+				if statusFilter == 0 && ch.Status == common.ChannelStatusEnabled {
+					continue
+				}
+				if typeFilter >= 0 && ch.Type != typeFilter {
+					continue
+				}
+				filtered = append(filtered, ch)
+			}
+			channelData = append(channelData, filtered...)
 		}
-		// 计算 tag 总数用于分页
 		total, _ = model.CountAllTags()
-	} else if typeFilter >= 0 {
-		channels, err := model.GetChannelsByType((p-1)*pageSize, pageSize, idSort, typeFilter)
-		if err != nil {
-			c.JSON(http.StatusOK, gin.H{"success": false, "message": err.Error()})
-			return
-		}
-		channelData = channels
-		total, _ = model.CountChannelsByType(typeFilter)
 	} else {
-		channels, err := model.GetAllChannels((p-1)*pageSize, pageSize, false, idSort)
+		baseQuery := model.DB.Model(&model.Channel{})
+		if typeFilter >= 0 {
+			baseQuery = baseQuery.Where("type = ?", typeFilter)
+		}
+		if statusFilter == common.ChannelStatusEnabled {
+			baseQuery = baseQuery.Where("status = ?", common.ChannelStatusEnabled)
+		} else if statusFilter == 0 {
+			baseQuery = baseQuery.Where("status != ?", common.ChannelStatusEnabled)
+		}
+
+		baseQuery.Count(&total)
+
+		order := "priority desc"
+		if idSort {
+			order = "id desc"
+		}
+
+		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
 		}
-		channelData = channels
-		total, _ = model.CountAllChannels()
 	}
 
-	// calculate type counts
-	typeCounts, _ := model.CountChannelsGroupByType()
+	countQuery := model.DB.Model(&model.Channel{})
+	if statusFilter == common.ChannelStatusEnabled {
+		countQuery = countQuery.Where("status = ?", common.ChannelStatusEnabled)
+	} else if statusFilter == 0 {
+		countQuery = countQuery.Where("status != ?", common.ChannelStatusEnabled)
+	}
+	var results []struct {
+		Type  int64
+		Count int64
+	}
+	_ = countQuery.Select("type, count(*) as count").Group("type").Find(&results).Error
+	typeCounts := make(map[int64]int64)
+	for _, r := range results {
+		typeCounts[r.Type] = r.Count
+	}
 
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
@@ -134,22 +182,15 @@ func FetchUpstreamModels(c *gin.Context) {
 		return
 	}
 
-	//if channel.Type != common.ChannelTypeOpenAI {
-	//	c.JSON(http.StatusOK, gin.H{
-	//		"success": false,
-	//		"message": "仅支持 OpenAI 类型渠道",
-	//	})
-	//	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))
@@ -173,7 +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)
@@ -206,6 +247,8 @@ func SearchChannels(c *gin.Context) {
 	keyword := c.Query("keyword")
 	group := c.Query("group")
 	modelKeyword := c.Query("model")
+	statusParam := c.Query("status")
+	statusFilter := parseStatusFilter(statusParam)
 	idSort, _ := strconv.ParseBool(c.Query("id_sort"))
 	enableTagMode, _ := strconv.ParseBool(c.Query("tag_mode"))
 	channelData := make([]*model.Channel, 0)
@@ -238,17 +281,71 @@ func SearchChannels(c *gin.Context) {
 		channelData = channels
 	}
 
+	if statusFilter == common.ChannelStatusEnabled || statusFilter == 0 {
+		filtered := make([]*model.Channel, 0, len(channelData))
+		for _, ch := range channelData {
+			if statusFilter == common.ChannelStatusEnabled && ch.Status != common.ChannelStatusEnabled {
+				continue
+			}
+			if statusFilter == 0 && ch.Status == common.ChannelStatusEnabled {
+				continue
+			}
+			filtered = append(filtered, ch)
+		}
+		channelData = filtered
+	}
+
 	// calculate type counts for search results
 	typeCounts := make(map[int64]int64)
 	for _, channel := range channelData {
 		typeCounts[int64(channel.Type)]++
 	}
 
+	typeParam := c.Query("type")
+	typeFilter := -1
+	if typeParam != "" {
+		if tp, err := strconv.Atoi(typeParam); err == nil {
+			typeFilter = tp
+		}
+	}
+
+	if typeFilter >= 0 {
+		filtered := make([]*model.Channel, 0, len(channelData))
+		for _, ch := range channelData {
+			if ch.Type == typeFilter {
+				filtered = append(filtered, ch)
+			}
+		}
+		channelData = filtered
+	}
+
+	page, _ := strconv.Atoi(c.DefaultQuery("p", "1"))
+	pageSize, _ := strconv.Atoi(c.DefaultQuery("page_size", "20"))
+	if page < 1 {
+		page = 1
+	}
+	if pageSize <= 0 {
+		pageSize = 20
+	}
+
+	total := len(channelData)
+	startIdx := (page - 1) * pageSize
+	if startIdx > total {
+		startIdx = total
+	}
+	endIdx := startIdx + pageSize
+	if endIdx > total {
+		endIdx = total
+	}
+
+	pagedData := channelData[startIdx:endIdx]
+
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
 		"data": gin.H{
-			"items":       channelData,
+			"items":       pagedData,
+			"total":       total,
 			"type_counts": typeCounts,
 		},
 	})
@@ -292,7 +389,7 @@ func AddChannel(c *gin.Context) {
 	}
 	channel.CreatedTime = common.GetTimestamp()
 	keys := strings.Split(channel.Key, "\n")
-	if channel.Type == common.ChannelTypeVertexAi {
+	if channel.Type == constant.ChannelTypeVertexAi {
 		if channel.Other == "" {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
@@ -517,7 +614,7 @@ func UpdateChannel(c *gin.Context) {
 		})
 		return
 	}
-	if channel.Type == common.ChannelTypeVertexAi {
+	if channel.Type == constant.ChannelTypeVertexAi {
 		if channel.Other == "" {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
@@ -546,6 +643,7 @@ func UpdateChannel(c *gin.Context) {
 		})
 		return
 	}
+	channel.Key = ""
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
@@ -571,7 +669,7 @@ 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)
@@ -156,17 +115,16 @@ 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),
 				})
 			}
 		}
@@ -181,14 +139,14 @@ func ListModels(c *gin.Context) {
 			return
 		}
 		group := userGroup
-		tokenGroup := c.GetString("token_group")
+		tokenGroup := common.GetContextKeyString(c, constant.ContextKeyUserGroup)
 		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

@@ -65,7 +65,7 @@ func Playground(c *gin.Context) {
 		return
 	}
 	middleware.SetupContextForSelectedChannel(c, channel, playgroundRequest.Model)
-	c.Set(constant.ContextKeyRequestStartTime, time.Now())
+	common.SetContextKey(c, constant.ContextKeyRequestStartTime, time.Now())
 
 	// Write user context to ensure acceptUnsetRatio is available
 	userId := c.GetInt("id")

controller/ratio_sync.go

@@ -3,6 +3,7 @@ package controller
 import (
     "context"
     "encoding/json"
+    "fmt"
     "net/http"
     "strings"
     "sync"
@@ -43,7 +44,17 @@ func FetchUpstreamRatios(c *gin.Context) {
 
     var upstreams []dto.UpstreamDTO
 
-    if len(req.ChannelIDs) > 0 {
+    if len(req.Upstreams) > 0 {
+        for _, u := range req.Upstreams {
+            if strings.HasPrefix(u.BaseURL, "http") {
+                if u.Endpoint == "" {
+                    u.Endpoint = defaultEndpoint
+                }
+                u.BaseURL = strings.TrimRight(u.BaseURL, "/")
+                upstreams = append(upstreams, u)
+            }
+        }
+    } else if len(req.ChannelIDs) > 0 {
         intIds := make([]int, 0, len(req.ChannelIDs))
         for _, id64 := range req.ChannelIDs {
             intIds = append(intIds, int(id64))
@@ -57,6 +68,7 @@ func FetchUpstreamRatios(c *gin.Context) {
         for _, ch := range dbChannels {
             if base := ch.GetBaseURL(); strings.HasPrefix(base, "http") {
                 upstreams = append(upstreams, dto.UpstreamDTO{
+                    ID:       ch.Id,
                     Name:     ch.Name,
                     BaseURL:  strings.TrimRight(base, "/"),
                     Endpoint: "",
@@ -93,43 +105,125 @@ func FetchUpstreamRatios(c *gin.Context) {
             }
             fullURL := chItem.BaseURL + endpoint
 
+            uniqueName := chItem.Name
+            if chItem.ID != 0 {
+                uniqueName = fmt.Sprintf("%s(%d)", chItem.Name, chItem.ID)
+            }
+
             ctx, cancel := context.WithTimeout(c.Request.Context(), time.Duration(req.Timeout)*time.Second)
             defer cancel()
 
             httpReq, err := http.NewRequestWithContext(ctx, http.MethodGet, fullURL, nil)
             if err != nil {
                 common.LogWarn(c.Request.Context(), "build request failed: "+err.Error())
-                ch <- upstreamResult{Name: chItem.Name, Err: err.Error()}
+                ch <- upstreamResult{Name: uniqueName, Err: err.Error()}
                 return
             }
 
             resp, err := client.Do(httpReq)
             if err != nil {
                 common.LogWarn(c.Request.Context(), "http error on "+chItem.Name+": "+err.Error())
-                ch <- upstreamResult{Name: chItem.Name, Err: err.Error()}
+                ch <- upstreamResult{Name: uniqueName, Err: err.Error()}
                 return
             }
             defer resp.Body.Close()
             if resp.StatusCode != http.StatusOK {
                 common.LogWarn(c.Request.Context(), "non-200 from "+chItem.Name+": "+resp.Status)
-                ch <- upstreamResult{Name: chItem.Name, Err: resp.Status}
+                ch <- upstreamResult{Name: uniqueName, Err: resp.Status}
                 return
             }
+            // 兼容两种上游接口格式：
+            //  type1: /api/ratio_config -> data 为 map[string]any，包含 model_ratio/completion_ratio/cache_ratio/model_price
+            //  type2: /api/pricing      -> data 为 []Pricing 列表，需要转换为与 type1 相同的 map 格式
             var body struct {
-                Success bool                   `json:"success"`
-                Data    map[string]any         `json:"data"`
-                Message string                 `json:"message"`
+                Success bool            `json:"success"`
+                Data    json.RawMessage `json:"data"`
+                Message string          `json:"message"`
             }
+
             if err := json.NewDecoder(resp.Body).Decode(&body); err != nil {
                 common.LogWarn(c.Request.Context(), "json decode failed from "+chItem.Name+": "+err.Error())
-                ch <- upstreamResult{Name: chItem.Name, Err: err.Error()}
+                ch <- upstreamResult{Name: uniqueName, Err: err.Error()}
                 return
             }
+
             if !body.Success {
-                ch <- upstreamResult{Name: chItem.Name, Err: body.Message}
+                ch <- upstreamResult{Name: uniqueName, Err: body.Message}
                 return
             }
-            ch <- upstreamResult{Name: chItem.Name, Data: body.Data}
+
+            // 尝试按 type1 解析
+            var type1Data map[string]any
+            if err := json.Unmarshal(body.Data, &type1Data); err == nil {
+                // 如果包含至少一个 ratioTypes 字段，则认为是 type1
+                isType1 := false
+                for _, rt := range ratioTypes {
+                    if _, ok := type1Data[rt]; ok {
+                        isType1 = true
+                        break
+                    }
+                }
+                if isType1 {
+                    ch <- upstreamResult{Name: uniqueName, Data: type1Data}
+                    return
+                }
+            }
+
+            // 如果不是 type1，则尝试按 type2 (/api/pricing) 解析
+            var pricingItems []struct {
+                ModelName       string  `json:"model_name"`
+                QuotaType       int     `json:"quota_type"`
+                ModelRatio      float64 `json:"model_ratio"`
+                ModelPrice      float64 `json:"model_price"`
+                CompletionRatio float64 `json:"completion_ratio"`
+            }
+            if err := json.Unmarshal(body.Data, &pricingItems); err != nil {
+                common.LogWarn(c.Request.Context(), "unrecognized data format from "+chItem.Name+": "+err.Error())
+                ch <- upstreamResult{Name: uniqueName, Err: "无法解析上游返回数据"}
+                return
+            }
+
+            modelRatioMap := make(map[string]float64)
+            completionRatioMap := make(map[string]float64)
+            modelPriceMap := make(map[string]float64)
+
+            for _, item := range pricingItems {
+                if item.QuotaType == 1 {
+                    modelPriceMap[item.ModelName] = item.ModelPrice
+                } else {
+                    modelRatioMap[item.ModelName] = item.ModelRatio
+                    // completionRatio 可能为 0，此时也直接赋值，保持与上游一致
+                    completionRatioMap[item.ModelName] = item.CompletionRatio
+                }
+            }
+
+            converted := make(map[string]any)
+
+            if len(modelRatioMap) > 0 {
+                ratioAny := make(map[string]any, len(modelRatioMap))
+                for k, v := range modelRatioMap {
+                    ratioAny[k] = v
+                }
+                converted["model_ratio"] = ratioAny
+            }
+
+            if len(completionRatioMap) > 0 {
+                compAny := make(map[string]any, len(completionRatioMap))
+                for k, v := range completionRatioMap {
+                    compAny[k] = v
+                }
+                converted["completion_ratio"] = compAny
+            }
+
+            if len(modelPriceMap) > 0 {
+                priceAny := make(map[string]any, len(modelPriceMap))
+                for k, v := range modelPriceMap {
+                    priceAny[k] = v
+                }
+                converted["model_price"] = priceAny
+            }
+
+            ch <- upstreamResult{Name: uniqueName, Data: converted}
         }(chn)
     }
 
@@ -202,6 +296,43 @@ func buildDifferences(localData map[string]any, successfulChannels []struct {
         }
     }
 
+    confidenceMap := make(map[string]map[string]bool)
+    
+    // 预处理阶段：检查pricing接口的可信度
+    for _, channel := range successfulChannels {
+        confidenceMap[channel.name] = make(map[string]bool)
+        
+        modelRatios, hasModelRatio := channel.data["model_ratio"].(map[string]any)
+        completionRatios, hasCompletionRatio := channel.data["completion_ratio"].(map[string]any)
+        
+        if hasModelRatio && hasCompletionRatio {
+            // 遍历所有模型，检查是否满足不可信条件
+            for modelName := range allModels {
+                // 默认为可信
+                confidenceMap[channel.name][modelName] = true
+                
+                // 检查是否满足不可信条件：model_ratio为37.5且completion_ratio为1
+                if modelRatioVal, ok := modelRatios[modelName]; ok {
+                    if completionRatioVal, ok := completionRatios[modelName]; ok {
+                        // 转换为float64进行比较
+                        if modelRatioFloat, ok := modelRatioVal.(float64); ok {
+                            if completionRatioFloat, ok := completionRatioVal.(float64); ok {
+                                if modelRatioFloat == 37.5 && completionRatioFloat == 1.0 {
+                                    confidenceMap[channel.name][modelName] = false
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        } else {
+            // 如果不是从pricing接口获取的数据，则全部标记为可信
+            for modelName := range allModels {
+                confidenceMap[channel.name][modelName] = true
+            }
+        }
+    }
+
     for modelName := range allModels {
         for _, ratioType := range ratioTypes {
             var localValue interface{} = nil
@@ -214,6 +345,7 @@ func buildDifferences(localData map[string]any, successfulChannels []struct {
             }
 
             upstreamValues := make(map[string]interface{})
+            confidenceValues := make(map[string]bool)
             hasUpstreamValue := false
             hasDifference := false
 
@@ -241,6 +373,8 @@ func buildDifferences(localData map[string]any, successfulChannels []struct {
                 }
                 
                 upstreamValues[channel.name] = upstreamValue
+                
+                confidenceValues[channel.name] = confidenceMap[channel.name][modelName]
             }
 
             shouldInclude := false
@@ -262,6 +396,7 @@ func buildDifferences(localData map[string]any, successfulChannels []struct {
                 differences[modelName][ratioType] = dto.DifferenceItem{
                     Current:   localValue,
                     Upstreams: upstreamValues,
+                    Confidence: confidenceValues,
                 }
             }
         }
@@ -283,9 +418,26 @@ func buildDifferences(localData map[string]any, successfulChannels []struct {
             for chName := range item.Upstreams {
                 if !channelHasDiff[chName] {
                     delete(item.Upstreams, chName)
+                    delete(item.Confidence, chName)
                 }
             }
-            differences[modelName][ratioType] = item
+
+            allSame := true
+            for _, v := range item.Upstreams {
+                if v != "same" {
+                    allSame = false
+                    break
+                }
+            }
+            if len(item.Upstreams) == 0 || allSame {
+                delete(ratioMap, ratioType)
+            } else {
+                differences[modelName][ratioType] = item
+            }
+        }
+
+        if len(ratioMap) == 0 {
+            delete(differences, modelName)
         }
     }
 

controller/relay.go

@@ -8,12 +8,12 @@ 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"
@@ -69,7 +69,7 @@ func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode
 }
 
 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")
@@ -132,7 +132,7 @@ func WssRelay(c *gin.Context) {
 		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
@@ -295,7 +295,7 @@ 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

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/token.go

@@ -258,3 +258,32 @@ func UpdateToken(c *gin.Context) {
 	})
 	return
 }
+
+type TokenBatch struct {
+	Ids []int `json:"ids"`
+}
+
+func DeleteTokenBatch(c *gin.Context) {
+	tokenBatch := TokenBatch{}
+	if err := c.ShouldBindJSON(&tokenBatch); err != nil || len(tokenBatch.Ids) == 0 {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "参数错误",
+		})
+		return
+	}
+	userId := c.GetInt("id")
+	count, err := model.BatchDeleteTokens(tokenBatch.Ids, userId)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "",
+		"data":    count,
+	})
+}

controller/user.go

@@ -246,15 +246,15 @@ 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 +262,13 @@ 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 +487,7 @@ 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)
 			}

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/dalle.go

@@ -15,6 +15,7 @@ type ImageRequest struct {
 	Background     string          `json:"background,omitempty"`
 	Moderation     string          `json:"moderation,omitempty"`
 	OutputFormat   string          `json:"output_format,omitempty"`
+	Watermark      *bool           `json:"watermark,omitempty"`
 }
 
 type ImageResponse struct {

dto/openai_request.go

@@ -57,6 +57,7 @@ type GeneralOpenAIRequest struct {
 	ExtraBody           json.RawMessage   `json:"extra_body,omitempty"`
 	WebSearchOptions    *WebSearchOptions `json:"web_search_options,omitempty"`
 	// OpenRouter Params
+	Usage     json.RawMessage `json:"usage,omitempty"`
 	Reasoning json.RawMessage `json:"reasoning,omitempty"`
 	// Ali Qwen Params
 	VlHighResolutionImages json.RawMessage `json:"vl_high_resolution_images,omitempty"`
@@ -65,7 +66,7 @@ type GeneralOpenAIRequest struct {
 func (r *GeneralOpenAIRequest) ToMap() map[string]any {
 	result := make(map[string]any)
 	data, _ := common.EncodeJson(r)
-	_ = common.DecodeJson(data, &result)
+	_ = common.UnmarshalJson(data, &result)
 	return result
 }
 
@@ -645,4 +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

@@ -26,7 +26,7 @@ type OpenAITextResponse struct {
 	Id      string                     `json:"id"`
 	Model   string                     `json:"model"`
 	Object  string                     `json:"object"`
-	Created int64                      `json:"created"`
+	Created any                        `json:"created"`
 	Choices []OpenAITextResponseChoice `json:"choices"`
 	Error   *OpenAIError               `json:"error,omitempty"`
 	Usage   `json:"usage"`
@@ -178,6 +178,8 @@ type Usage struct {
 	InputTokens            int                `json:"input_tokens"`
 	OutputTokens           int                `json:"output_tokens"`
 	InputTokensDetails     *InputTokenDetails `json:"input_tokens_details"`
+	// OpenRouter Params
+	Cost float64 `json:"cost,omitempty"`
 }
 
 type InputTokenDetails struct {

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/ratio_sync.go

@@ -1,18 +1,7 @@
 package dto
 
-// UpstreamDTO 提交到后端同步倍率的上游渠道信息
-// Endpoint 可以为空，后端会默认使用 /api/ratio_config
-// BaseURL 必须以 http/https 开头，不要以 / 结尾
-// 例如： https://api.example.com
-// Endpoint: /api/ratio_config
-// 提交示例：
-// {
-//   "name": "openai",
-//   "base_url": "https://api.openai.com",
-//   "endpoint": "/ratio_config"
-// }
-
 type UpstreamDTO struct {
+    ID       int    `json:"id,omitempty"`
     Name     string `json:"name" binding:"required"`
     BaseURL  string `json:"base_url" binding:"required"`
     Endpoint string `json:"endpoint"`
@@ -20,6 +9,7 @@ type UpstreamDTO struct {
 
 type UpstreamRequest struct {
     ChannelIDs []int64 `json:"channel_ids"`
+    Upstreams   []UpstreamDTO `json:"upstreams"`
     Timeout    int     `json:"timeout"`
 }
 
@@ -37,10 +27,9 @@ type TestResult struct {
 type DifferenceItem struct {
     Current   interface{}            `json:"current"`
     Upstreams map[string]interface{} `json:"upstreams"`
+    Confidence map[string]bool       `json:"confidence"`
 }
 
-// SyncableChannel 可同步的渠道信息（base_url 不为空）
-
 type SyncableChannel struct {
     ID      int    `json:"id"`
     Name    string `json:"name"`

main.go

@@ -32,12 +32,12 @@ var buildFS embed.FS
 var indexPage []byte
 
 func main() {
-	err := godotenv.Load(".env")
-	if err != nil {
-		common.SysLog("Support for .env file is disabled: " + err.Error())
-	}
 
-	common.LoadEnv()
+	err := InitResources()
+	if err != nil {
+		common.FatalLog("failed to initialize resources: " + err.Error())
+		return
+	}
 
 	common.SetupLogger()
 	common.SysLog("New API " + common.Version + " started")
@@ -47,19 +47,7 @@ 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 +55,6 @@ 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
@@ -186,3 +159,51 @@ 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.")
+	}
+
+	// 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

@@ -184,7 +184,7 @@ 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)

middleware/distributor.go

@@ -25,7 +25,7 @@ type ModelRequest struct {
 
 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 +34,14 @@ 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 +57,7 @@ func Distribute() func(c *gin.Context) {
 			}
 			userGroup = tokenGroup
 		}
-		c.Set("group", userGroup)
+		common.SetContextKey(c, constant.ContextKeyUsingGroup, userGroup)
 		if ok {
 			id, err := strconv.Atoi(channelId.(string))
 			if err != nil {
@@ -76,9 +76,9 @@ 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 +121,7 @@ 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 +171,25 @@ 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)
@@ -251,21 +261,21 @@ func SetupContextForSelectedChannel(c *gin.Context, channel *model.Channel, mode
 	c.Set("base_url", 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

@@ -21,7 +21,22 @@ 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 +61,7 @@ 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 +87,14 @@ 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)
 		}
 	}
 

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.go

@@ -327,3 +327,37 @@ func CountUserTokens(userId int) (int64, error) {
 	err := DB.Model(&Token{}).Where("user_id = ?", userId).Count(&total).Error
 	return total, err
 }
+
+// BatchDeleteTokens 删除指定用户的一组令牌，返回成功删除数量
+func BatchDeleteTokens(ids []int, userId int) (int, error) {
+	if len(ids) == 0 {
+		return 0, errors.New("ids 不能为空！")
+	}
+
+	tx := DB.Begin()
+
+	var tokens []Token
+	if err := tx.Where("user_id = ? AND id IN (?)", userId, ids).Find(&tokens).Error; err != nil {
+		tx.Rollback()
+		return 0, err
+	}
+
+	if err := tx.Where("user_id = ? AND id IN (?)", userId, ids).Delete(&Token{}).Error; err != nil {
+		tx.Rollback()
+		return 0, err
+	}
+
+	if err := tx.Commit().Error; err != nil {
+		return 0, err
+	}
+
+	if common.RedisEnabled {
+		gopool.Go(func() {
+			for _, t := range tokens {
+				_ = cacheDeleteToken(t.Key)
+			}
+		})
+	}
+
+	return len(tokens), nil
+}

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

@@ -114,7 +114,7 @@ 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 +134,7 @@ 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

model/user_cache.go

@@ -24,12 +24,12 @@ 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{} {
@@ -70,7 +70,7 @@ 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,
 	)
 }
 

relay/channel/adapter.go

@@ -45,5 +45,5 @@ 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

@@ -30,7 +30,7 @@ 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 +82,7 @@ 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) {

relay/channel/ali/image.go

@@ -132,10 +132,7 @@ func aliImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &aliTaskResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/ali/rerank.go

@@ -4,6 +4,7 @@ import (
 	"encoding/json"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
@@ -35,10 +36,7 @@ func RerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayI
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 
 	var aliResponse AliRerankResponse
 	err = json.Unmarshal(responseBody, &aliResponse)

relay/channel/ali/text.go

@@ -39,34 +39,18 @@ 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)
+	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
 	}
 
-	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
@@ -186,10 +170,7 @@ 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
 }
 
@@ -199,10 +180,7 @@ func aliHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStatus
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &aliResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/baidu/relay-baidu.go

@@ -166,10 +166,7 @@ func baiduStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWi
 			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
 }
 
@@ -179,10 +176,7 @@ func baiduHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStat
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
@@ -215,10 +209,7 @@ func baiduEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErro
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
@@ -280,7 +271,7 @@ 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/claude/relay-claude.go

@@ -7,6 +7,7 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
+	"one-api/relay/channel/openrouter"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/service"
@@ -122,6 +123,21 @@ func RequestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*dto.Cla
 		claudeRequest.Model = strings.TrimSuffix(textRequest.Model, "-thinking")
 	}
 
+	if textRequest.Reasoning != nil {
+		var reasoning openrouter.RequestReasoning
+		if err := common.UnmarshalJson(textRequest.Reasoning, &reasoning); err != nil {
+			return nil, err
+		}
+
+		budgetTokens := reasoning.MaxTokens
+		if budgetTokens > 0 {
+			claudeRequest.Thinking = &dto.Thinking{
+				Type:         "enabled",
+				BudgetTokens: &budgetTokens,
+			}
+		}
+	}
+
 	if textRequest.Stop != nil {
 		// stop maybe string/array string, convert to array string
 		switch textRequest.Stop.(type) {
@@ -503,7 +519,7 @@ func FormatClaudeResponseInfo(requestMode int, claudeResponse *dto.ClaudeRespons
 
 func HandleStreamResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data string, requestMode int) *dto.OpenAIErrorWithStatusCode {
 	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)
@@ -603,7 +619,7 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 
 func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claudeInfo *ClaudeResponseInfo, data []byte, requestMode int) *dto.OpenAIErrorWithStatusCode {
 	var claudeResponse dto.ClaudeResponse
-	err := common.DecodeJson(data, &claudeResponse)
+	err := common.UnmarshalJson(data, &claudeResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_claude_response_failed", http.StatusInternalServerError)
 	}
@@ -641,13 +657,14 @@ func HandleClaudeResponseData(c *gin.Context, info *relaycommon.RelayInfo, claud
 	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) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	claudeInfo := &ClaudeResponseInfo{
 		ResponseId:   fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
 		Created:      common.GetTimestamp(),
@@ -659,7 +676,6 @@ func ClaudeHandler(c *gin.Context, resp *http.Response, requestMode int, info *r
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
 	}
-	resp.Body.Close()
 	if common.DebugEnabled {
 		println("responseBody: ", string(responseBody))
 	}

relay/channel/cloudflare/relay_cloudflare.go

@@ -81,10 +81,7 @@ 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
 }
@@ -94,10 +91,7 @@ func cfHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo)
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	var response dto.TextResponse
 	err = json.Unmarshal(responseBody, &response)
 	if err != nil {
@@ -127,10 +121,7 @@ func cfSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayIn
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &cfResp)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/cohere/relay-cohere.go

@@ -173,10 +173,7 @@ func cohereHandler(c *gin.Context, resp *http.Response, modelName string, prompt
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	var cohereResp CohereResponseResult
 	err = json.Unmarshal(responseBody, &cohereResp)
 	if err != nil {
@@ -217,10 +214,7 @@ func cohereRerankHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	var cohereResp CohereRerankResponseResult
 	err = json.Unmarshal(responseBody, &cohereResp)
 	if err != nil {

relay/channel/coze/relay-coze.go

@@ -48,10 +48,7 @@ func cozeChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	// convert coze response to openai response
 	var response dto.TextResponse
 	var cozeResponse CozeChatDetailResponse

relay/channel/dify/relay-dify.go

@@ -95,7 +95,7 @@ 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())
@@ -257,10 +257,7 @@ func difyHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInf
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &difyResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

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

@@ -1,7 +1,6 @@
 package gemini
 
 import (
-	"encoding/json"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -15,15 +14,13 @@ import (
 )
 
 func GeminiTextGenerationHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.Usage, *dto.OpenAIErrorWithStatusCode) {
+	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)
-	}
 
 	if common.DebugEnabled {
 		println(string(responseBody))
@@ -31,7 +28,7 @@ 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.UnmarshalJson(responseBody, &geminiResponse)
 	if err != nil {
 		return nil, service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError)
 	}
@@ -54,18 +51,12 @@ 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.EncodeJson(geminiResponse)
 	if err != nil {
 		return nil, service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError)
 	}
 
-	// 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
 }
@@ -80,7 +71,7 @@ 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

@@ -78,32 +78,12 @@ func clampThinkingBudget(modelName string, budget int) int {
 	return budget
 }
 
-// Setting safety to the lowest possible values since Gemini is already powerless enough
-func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest, info *relaycommon.RelayInfo) (*GeminiChatRequest, error) {
-
-	geminiRequest := GeminiChatRequest{
-		Contents: make([]GeminiChatContent, 0, len(textRequest.Messages)),
-		GenerationConfig: GeminiChatGenerationConfig{
-			Temperature:     textRequest.Temperature,
-			TopP:            textRequest.TopP,
-			MaxOutputTokens: textRequest.MaxTokens,
-			Seed:            int64(textRequest.Seed),
-		},
-	}
-
-	if model_setting.IsGeminiModelSupportImagine(info.UpstreamModelName) {
-		geminiRequest.GenerationConfig.ResponseModalities = []string{
-			"TEXT",
-			"IMAGE",
-		}
-	}
-
+func ThinkingAdaptor(geminiRequest *GeminiChatRequest, info *relaycommon.RelayInfo) {
 	if model_setting.GetGeminiSettings().ThinkingAdapterEnabled {
 		modelName := info.UpstreamModelName
 		isNew25Pro := strings.HasPrefix(modelName, "gemini-2.5-pro") &&
 			!strings.HasPrefix(modelName, "gemini-2.5-pro-preview-05-06") &&
 			!strings.HasPrefix(modelName, "gemini-2.5-pro-preview-03-25")
-		is25FlashLite := strings.HasPrefix(modelName, "gemini-2.5-flash-lite")
 
 		if strings.Contains(modelName, "-thinking-") {
 			parts := strings.SplitN(modelName, "-thinking-", 2)
@@ -134,21 +114,46 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest, info *relaycommon
 					IncludeThoughts: true,
 				}
 			} else {
-				budgetTokens := model_setting.GetGeminiSettings().ThinkingAdapterBudgetTokensPercentage * float64(geminiRequest.GenerationConfig.MaxOutputTokens)
-				clampedBudget := clampThinkingBudget(modelName, int(budgetTokens))
 				geminiRequest.GenerationConfig.ThinkingConfig = &GeminiThinkingConfig{
-					ThinkingBudget:  common.GetPointer(clampedBudget),
 					IncludeThoughts: true,
 				}
+				if geminiRequest.GenerationConfig.MaxOutputTokens > 0 {
+					budgetTokens := model_setting.GetGeminiSettings().ThinkingAdapterBudgetTokensPercentage * float64(geminiRequest.GenerationConfig.MaxOutputTokens)
+					clampedBudget := clampThinkingBudget(modelName, int(budgetTokens))
+					geminiRequest.GenerationConfig.ThinkingConfig.ThinkingBudget = common.GetPointer(clampedBudget)
+				}
 			}
 		} else if strings.HasSuffix(modelName, "-nothinking") {
-			if !isNew25Pro && !is25FlashLite {
+			if !isNew25Pro {
 				geminiRequest.GenerationConfig.ThinkingConfig = &GeminiThinkingConfig{
 					ThinkingBudget: common.GetPointer(0),
 				}
 			}
 		}
 	}
+}
+
+// Setting safety to the lowest possible values since Gemini is already powerless enough
+func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest, info *relaycommon.RelayInfo) (*GeminiChatRequest, error) {
+
+	geminiRequest := GeminiChatRequest{
+		Contents: make([]GeminiChatContent, 0, len(textRequest.Messages)),
+		GenerationConfig: GeminiChatGenerationConfig{
+			Temperature:     textRequest.Temperature,
+			TopP:            textRequest.TopP,
+			MaxOutputTokens: textRequest.MaxTokens,
+			Seed:            int64(textRequest.Seed),
+		},
+	}
+
+	if model_setting.IsGeminiModelSupportImagine(info.UpstreamModelName) {
+		geminiRequest.GenerationConfig.ResponseModalities = []string{
+			"TEXT",
+			"IMAGE",
+		}
+	}
+
+	ThinkingAdaptor(&geminiRequest, info)
 
 	safetySettings := make([]GeminiChatSafetySettings, 0, len(SafetySettingList))
 	for _, category := range SafetySettingList {
@@ -796,7 +801,7 @@ 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
@@ -861,15 +866,12 @@ func GeminiChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	if common.DebugEnabled {
 		println(string(responseBody))
 	}
 	var geminiResponse GeminiChatResponse
-	err = common.DecodeJson(responseBody, &geminiResponse)
+	err = common.UnmarshalJson(responseBody, &geminiResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
@@ -915,11 +917,12 @@ func GeminiChatHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 }
 
 func GeminiEmbeddingHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	responseBody, readErr := io.ReadAll(resp.Body)
 	if readErr != nil {
 		return nil, service.OpenAIErrorWrapper(readErr, "read_response_body_failed", http.StatusInternalServerError)
 	}
-	_ = resp.Body.Close()
 
 	var geminiResponse GeminiEmbeddingResponse
 	if jsonErr := json.Unmarshal(responseBody, &geminiResponse); jsonErr != nil {
@@ -951,14 +954,11 @@ func GeminiEmbeddingHandler(c *gin.Context, resp *http.Response, info *relaycomm
 	}
 	openAIResponse.Usage = *usage.(*dto.Usage)
 
-	jsonResponse, jsonErr := json.Marshal(openAIResponse)
+	jsonResponse, jsonErr := common.EncodeJson(openAIResponse)
 	if jsonErr != nil {
 		return nil, service.OpenAIErrorWrapper(jsonErr, "marshal_response_failed", http.StatusInternalServerError)
 	}
 
-	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/mokaai/relay-mokaai.go

@@ -5,6 +5,7 @@ import (
 	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
 	"one-api/service"
 )
@@ -26,7 +27,7 @@ func embeddingRequestOpenAI2Moka(request dto.GeneralOpenAIRequest) *dto.Embeddin
 	}
 	return &dto.EmbeddingRequest{
 		Input: input,
-		Model:  request.Model,
+		Model: request.Model,
 	}
 }
 
@@ -53,10 +54,7 @@ func mokaEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &baiduResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
@@ -80,4 +78,3 @@ func mokaEmbeddingHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 	_, err = c.Writer.Write(jsonResponse)
 	return nil, &fullTextResponse.Usage
 }
-

relay/channel/ollama/relay-ollama.go

@@ -1,12 +1,12 @@
 package ollama
 
 import (
-	"bytes"
 	"encoding/json"
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
 	"one-api/service"
 	"strings"
@@ -88,10 +88,7 @@ func ollamaEmbeddingHandler(c *gin.Context, resp *http.Response, promptTokens in
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &ollamaEmbeddingResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
@@ -120,31 +117,7 @@ func ollamaEmbeddingHandler(c *gin.Context, resp *http.Response, promptTokens in
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
 	}
-	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
-	}
+	common.IOCopyBytesGracefully(c, resp, doResponseBody)
 	return nil, usage
 }
 

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"
@@ -21,7 +20,7 @@ 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"
 	"path/filepath"
 	"strings"
@@ -54,7 +53,7 @@ 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 think2Content, ok := info.ChannelSetting[constant.ChannelSettingThinkingToContent].(bool); ok && think2Content {
 		info.ThinkingContentInfo = relaycommon.ThinkingContentInfo{
 			IsFirstThinkingContent:  true,
 			SendLastThinkingContent: false,
@@ -67,7 +66,7 @@ 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 +78,10 @@ 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 +89,25 @@ 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 +118,14 @@ 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{
@@ -145,7 +144,7 @@ func (a *Adaptor) SetupRequestHeader(c *gin.Context, header *http.Header, info *
 	} else {
 		header.Set("Authorization", "Bearer "+info.ApiKey)
 	}
-	if info.ChannelType == common.ChannelTypeOpenRouter {
+	if info.ChannelType == constant.ChannelTypeOpenRouter {
 		header.Set("HTTP-Referer", "https://github.com/Calcium-Ion/new-api")
 		header.Set("X-Title", "New API")
 	}
@@ -156,9 +155,14 @@ 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 == constant.ChannelTypeOpenRouter {
+		if len(request.Usage) == 0 {
+			request.Usage = json.RawMessage(`{"include":true}`)
+		}
+	}
 	if strings.HasPrefix(request.Model, "o") {
 		if request.MaxCompletionTokens == 0 && request.MaxTokens != 0 {
 			request.MaxCompletionTokens = request.MaxTokens
@@ -200,7 +204,7 @@ 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)
@@ -249,7 +253,7 @@ 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)
@@ -406,11 +410,11 @@ 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)
@@ -419,19 +423,19 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
 	switch info.RelayMode {
-	case constant.RelayModeRealtime:
+	case relayconstant.RelayModeRealtime:
 		err, usage = OpenaiRealtimeHandler(c, info)
-	case constant.RelayModeAudioSpeech:
+	case relayconstant.RelayModeAudioSpeech:
 		err, usage = OpenaiTTSHandler(c, resp, info)
-	case constant.RelayModeAudioTranslation:
+	case relayconstant.RelayModeAudioTranslation:
 		fallthrough
-	case constant.RelayModeAudioTranscription:
+	case relayconstant.RelayModeAudioTranscription:
 		err, usage = OpenaiSTTHandler(c, resp, info, a.ResponseFormat)
-	case constant.RelayModeImagesGenerations, constant.RelayModeImagesEdits:
+	case relayconstant.RelayModeImagesGenerations, relayconstant.RelayModeImagesEdits:
 		err, usage = OpenaiHandlerWithUsage(c, resp, info)
-	case constant.RelayModeRerank:
+	case relayconstant.RelayModeRerank:
 		err, usage = common_handler.RerankHandler(c, info, resp)
-	case constant.RelayModeResponses:
+	case relayconstant.RelayModeResponses:
 		if info.IsStream {
 			err, usage = OaiResponsesStreamHandler(c, resp, info)
 		} else {
@@ -449,17 +453,17 @@ 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
@@ -468,17 +472,17 @@ 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"
@@ -34,7 +33,7 @@ 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
 	}
 
@@ -111,12 +110,13 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 		return service.OpenAIErrorWrapper(fmt.Errorf("invalid response"), "invalid_response", http.StatusInternalServerError), nil
 	}
 
-	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{}
@@ -148,31 +148,15 @@ 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 +168,7 @@ 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
 			}
@@ -197,16 +181,14 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 }
 
 func OpenaiHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	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
 	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = common.DecodeJson(responseBody, &simpleResponse)
+	err = common.UnmarshalJson(responseBody, &simpleResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
@@ -238,7 +220,7 @@ 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.EncodeJson(simpleResponse)
 			if err != nil {
 				return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
 			}
@@ -247,29 +229,15 @@ func OpenaiHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayI
 		}
 	case relaycommon.RelayFormatClaude:
 		claudeResp := service.ResponseOpenAI2Claude(&simpleResponse, info)
-		claudeRespStr, err := json.Marshal(claudeResp)
+		claudeRespStr, err := common.EncodeJson(claudeResp)
 		if err != nil {
 			return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
 		}
 		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()
+	common.IOCopyBytesGracefully(c, resp, responseBody)
+
 	return nil, &simpleResponse.Usage
 }
 
@@ -280,7 +248,7 @@ func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	// 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
@@ -297,6 +265,8 @@ func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 }
 
 func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, responseFormat string) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	// count tokens by audio file duration
 	audioTokens, err := countAudioTokens(c)
 	if err != nil {
@@ -306,25 +276,8 @@ func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	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])
-	}
-	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
@@ -415,7 +368,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 				}
 
 				realtimeEvent := &dto.RealtimeEvent{}
-				err = json.Unmarshal(message, realtimeEvent)
+				err = common.UnmarshalJson(message, realtimeEvent)
 				if err != nil {
 					errChan <- fmt.Errorf("error unmarshalling message: %v", err)
 					return
@@ -475,7 +428,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 				}
 				info.SetFirstResponseTime()
 				realtimeEvent := &dto.RealtimeEvent{}
-				err = json.Unmarshal(message, realtimeEvent)
+				err = common.UnmarshalJson(message, realtimeEvent)
 				if err != nil {
 					errChan <- fmt.Errorf("error unmarshalling message: %v", err)
 					return
@@ -522,9 +475,9 @@ 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
@@ -601,40 +554,25 @@ func preConsumeUsage(ctx *gin.Context, info *relaycommon.RelayInfo, usage *dto.R
 }
 
 func OpenaiHandlerWithUsage(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	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
-	}
 
 	var usageResp dto.SimpleResponse
-	err = json.Unmarshal(responseBody, &usageResp)
+	err = common.UnmarshalJson(responseBody, &usageResp)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "parse_response_body_failed", http.StatusInternalServerError), nil
 	}
+
+	// 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

relay/channel/openai/relay_responses.go

@@ -1,7 +1,6 @@
 package openai
 
 import (
-	"bytes"
 	"fmt"
 	"io"
 	"net/http"
@@ -16,17 +15,15 @@ import (
 )
 
 func OaiResponsesHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	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
 	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
-	err = common.DecodeJson(responseBody, &responsesResponse)
+	err = common.UnmarshalJson(responseBody, &responsesResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
@@ -41,22 +38,9 @@ func OaiResponsesHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 		}, 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)
-	// 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
@@ -82,7 +66,7 @@ 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":

relay/channel/palm/relay-palm.go

@@ -83,12 +83,7 @@ func palmStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWit
 			stopChan <- true
 			return
 		}
-		err = resp.Body.Close()
-		if err != nil {
-			common.SysError("error closing stream response: " + err.Error())
-			stopChan <- true
-			return
-		}
+		common.CloseResponseBodyGracefully(resp)
 		var palmResponse PaLMChatResponse
 		err = json.Unmarshal(responseBody, &palmResponse)
 		if err != nil {
@@ -122,10 +117,7 @@ func palmStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWit
 			return false
 		}
 	})
-	err := resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
-	}
+	common.CloseResponseBodyGracefully(resp)
 	return nil, responseText
 }
 
@@ -134,10 +126,7 @@ func palmHandler(c *gin.Context, resp *http.Response, promptTokens int, model st
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	var palmResponse PaLMChatResponse
 	err = json.Unmarshal(responseBody, &palmResponse)
 	if err != nil {

relay/channel/siliconflow/relay-siliconflow.go

@@ -5,6 +5,7 @@ import (
 	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/dto"
 	"one-api/service"
 )
@@ -14,10 +15,7 @@ func siliconflowRerankHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIE
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	var siliconflowResp SFRerankResponse
 	err = json.Unmarshal(responseBody, &siliconflowResp)
 	if err != nil {

relay/channel/task/jimeng/adaptor.go

@@ -0,0 +1,380 @@
+package jimeng
+
+import (
+	"bytes"
+	"crypto/hmac"
+	"crypto/sha256"
+	"encoding/hex"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"one-api/model"
+	"sort"
+	"strings"
+	"time"
+
+	"github.com/gin-gonic/gin"
+	"github.com/pkg/errors"
+
+	"one-api/common"
+	"one-api/constant"
+	"one-api/dto"
+	"one-api/relay/channel"
+	relaycommon "one-api/relay/common"
+	"one-api/service"
+)
+
+// ============================
+// Request / Response structures
+// ============================
+
+type requestPayload struct {
+	ReqKey           string   `json:"req_key"`
+	BinaryDataBase64 []string `json:"binary_data_base64,omitempty"`
+	ImageUrls        []string `json:"image_urls,omitempty"`
+	Prompt           string   `json:"prompt,omitempty"`
+	Seed             int64    `json:"seed"`
+	AspectRatio      string   `json:"aspect_ratio"`
+}
+
+type responsePayload struct {
+	Code      int    `json:"code"`
+	Message   string `json:"message"`
+	RequestId string `json:"request_id"`
+	Data      struct {
+		TaskID string `json:"task_id"`
+	} `json:"data"`
+}
+
+type responseTask struct {
+	Code int `json:"code"`
+	Data struct {
+		BinaryDataBase64 []interface{} `json:"binary_data_base64"`
+		ImageUrls        interface{}   `json:"image_urls"`
+		RespData         string        `json:"resp_data"`
+		Status           string        `json:"status"`
+		VideoUrl         string        `json:"video_url"`
+	} `json:"data"`
+	Message     string `json:"message"`
+	RequestId   string `json:"request_id"`
+	Status      int    `json:"status"`
+	TimeElapsed string `json:"time_elapsed"`
+}
+
+// ============================
+// Adaptor implementation
+// ============================
+
+type TaskAdaptor struct {
+	ChannelType int
+	accessKey   string
+	secretKey   string
+	baseURL     string
+}
+
+func (a *TaskAdaptor) Init(info *relaycommon.TaskRelayInfo) {
+	a.ChannelType = info.ChannelType
+	a.baseURL = info.BaseUrl
+
+	// apiKey format: "access_key|secret_key"
+	keyParts := strings.Split(info.ApiKey, "|")
+	if len(keyParts) == 2 {
+		a.accessKey = strings.TrimSpace(keyParts[0])
+		a.secretKey = strings.TrimSpace(keyParts[1])
+	}
+}
+
+// ValidateRequestAndSetAction parses body, validates fields and sets default action.
+func (a *TaskAdaptor) ValidateRequestAndSetAction(c *gin.Context, info *relaycommon.TaskRelayInfo) (taskErr *dto.TaskError) {
+	// Accept only POST /v1/video/generations as "generate" action.
+	action := constant.TaskActionGenerate
+	info.Action = action
+
+	req := relaycommon.TaskSubmitReq{}
+	if err := common.UnmarshalBodyReusable(c, &req); err != nil {
+		taskErr = service.TaskErrorWrapperLocal(err, "invalid_request", http.StatusBadRequest)
+		return
+	}
+	if strings.TrimSpace(req.Prompt) == "" {
+		taskErr = service.TaskErrorWrapperLocal(fmt.Errorf("prompt is required"), "invalid_request", http.StatusBadRequest)
+		return
+	}
+
+	// Store into context for later usage
+	c.Set("task_request", req)
+	return nil
+}
+
+// BuildRequestURL constructs the upstream URL.
+func (a *TaskAdaptor) BuildRequestURL(info *relaycommon.TaskRelayInfo) (string, error) {
+	return fmt.Sprintf("%s/?Action=CVSync2AsyncSubmitTask&Version=2022-08-31", a.baseURL), nil
+}
+
+// BuildRequestHeader sets required headers.
+func (a *TaskAdaptor) BuildRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.TaskRelayInfo) error {
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Accept", "application/json")
+	return a.signRequest(req, a.accessKey, a.secretKey)
+}
+
+// BuildRequestBody converts request into Jimeng specific format.
+func (a *TaskAdaptor) BuildRequestBody(c *gin.Context, info *relaycommon.TaskRelayInfo) (io.Reader, error) {
+	v, exists := c.Get("task_request")
+	if !exists {
+		return nil, fmt.Errorf("request not found in context")
+	}
+	req := v.(relaycommon.TaskSubmitReq)
+
+	body, err := a.convertToRequestPayload(&req)
+	if err != nil {
+		return nil, errors.Wrap(err, "convert request payload failed")
+	}
+	data, err := json.Marshal(body)
+	if err != nil {
+		return nil, err
+	}
+	return bytes.NewReader(data), nil
+}
+
+// DoRequest delegates to common helper.
+func (a *TaskAdaptor) DoRequest(c *gin.Context, info *relaycommon.TaskRelayInfo, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoTaskApiRequest(a, c, info, requestBody)
+}
+
+// DoResponse handles upstream response, returns taskID etc.
+func (a *TaskAdaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.TaskRelayInfo) (taskID string, taskData []byte, taskErr *dto.TaskError) {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		taskErr = service.TaskErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError)
+		return
+	}
+	_ = resp.Body.Close()
+
+	// Parse Jimeng response
+	var jResp responsePayload
+	if err := json.Unmarshal(responseBody, &jResp); err != nil {
+		taskErr = service.TaskErrorWrapper(errors.Wrapf(err, "body: %s", responseBody), "unmarshal_response_body_failed", http.StatusInternalServerError)
+		return
+	}
+
+	if jResp.Code != 10000 {
+		taskErr = service.TaskErrorWrapper(fmt.Errorf(jResp.Message), fmt.Sprintf("%d", jResp.Code), http.StatusInternalServerError)
+		return
+	}
+
+	c.JSON(http.StatusOK, gin.H{"task_id": jResp.Data.TaskID})
+	return jResp.Data.TaskID, responseBody, nil
+}
+
+// FetchTask fetch task status
+func (a *TaskAdaptor) FetchTask(baseUrl, key string, body map[string]any) (*http.Response, error) {
+	taskID, ok := body["task_id"].(string)
+	if !ok {
+		return nil, fmt.Errorf("invalid task_id")
+	}
+
+	uri := fmt.Sprintf("%s/?Action=CVSync2AsyncGetResult&Version=2022-08-31", baseUrl)
+	payload := map[string]string{
+		"req_key": "jimeng_vgfm_t2v_l20", // This is fixed value from doc: https://www.volcengine.com/docs/85621/1544774
+		"task_id": taskID,
+	}
+	payloadBytes, err := json.Marshal(payload)
+	if err != nil {
+		return nil, errors.Wrap(err, "marshal fetch task payload failed")
+	}
+
+	req, err := http.NewRequest(http.MethodPost, uri, bytes.NewBuffer(payloadBytes))
+	if err != nil {
+		return nil, err
+	}
+
+	req.Header.Set("Accept", "application/json")
+	req.Header.Set("Content-Type", "application/json")
+
+	keyParts := strings.Split(key, "|")
+	if len(keyParts) != 2 {
+		return nil, fmt.Errorf("invalid api key format for jimeng: expected 'ak|sk'")
+	}
+	accessKey := strings.TrimSpace(keyParts[0])
+	secretKey := strings.TrimSpace(keyParts[1])
+
+	if err := a.signRequest(req, accessKey, secretKey); err != nil {
+		return nil, errors.Wrap(err, "sign request failed")
+	}
+
+	return service.GetHttpClient().Do(req)
+}
+
+func (a *TaskAdaptor) GetModelList() []string {
+	return []string{"jimeng_vgfm_t2v_l20"}
+}
+
+func (a *TaskAdaptor) GetChannelName() string {
+	return "jimeng"
+}
+
+func (a *TaskAdaptor) signRequest(req *http.Request, accessKey, secretKey string) error {
+	var bodyBytes []byte
+	var err error
+
+	if req.Body != nil {
+		bodyBytes, err = io.ReadAll(req.Body)
+		if err != nil {
+			return errors.Wrap(err, "read request body failed")
+		}
+		_ = req.Body.Close()
+		req.Body = io.NopCloser(bytes.NewBuffer(bodyBytes)) // Rewind
+	} else {
+		bodyBytes = []byte{}
+	}
+
+	payloadHash := sha256.Sum256(bodyBytes)
+	hexPayloadHash := hex.EncodeToString(payloadHash[:])
+
+	t := time.Now().UTC()
+	xDate := t.Format("20060102T150405Z")
+	shortDate := t.Format("20060102")
+
+	req.Header.Set("Host", req.URL.Host)
+	req.Header.Set("X-Date", xDate)
+	req.Header.Set("X-Content-Sha256", hexPayloadHash)
+
+	// Sort and encode query parameters to create canonical query string
+	queryParams := req.URL.Query()
+	sortedKeys := make([]string, 0, len(queryParams))
+	for k := range queryParams {
+		sortedKeys = append(sortedKeys, k)
+	}
+	sort.Strings(sortedKeys)
+	var queryParts []string
+	for _, k := range sortedKeys {
+		values := queryParams[k]
+		sort.Strings(values)
+		for _, v := range values {
+			queryParts = append(queryParts, fmt.Sprintf("%s=%s", url.QueryEscape(k), url.QueryEscape(v)))
+		}
+	}
+	canonicalQueryString := strings.Join(queryParts, "&")
+
+	headersToSign := map[string]string{
+		"host":             req.URL.Host,
+		"x-date":           xDate,
+		"x-content-sha256": hexPayloadHash,
+	}
+	if req.Header.Get("Content-Type") != "" {
+		headersToSign["content-type"] = req.Header.Get("Content-Type")
+	}
+
+	var signedHeaderKeys []string
+	for k := range headersToSign {
+		signedHeaderKeys = append(signedHeaderKeys, k)
+	}
+	sort.Strings(signedHeaderKeys)
+
+	var canonicalHeaders strings.Builder
+	for _, k := range signedHeaderKeys {
+		canonicalHeaders.WriteString(k)
+		canonicalHeaders.WriteString(":")
+		canonicalHeaders.WriteString(strings.TrimSpace(headersToSign[k]))
+		canonicalHeaders.WriteString("\n")
+	}
+	signedHeaders := strings.Join(signedHeaderKeys, ";")
+
+	canonicalRequest := fmt.Sprintf("%s\n%s\n%s\n%s\n%s\n%s",
+		req.Method,
+		req.URL.Path,
+		canonicalQueryString,
+		canonicalHeaders.String(),
+		signedHeaders,
+		hexPayloadHash,
+	)
+
+	hashedCanonicalRequest := sha256.Sum256([]byte(canonicalRequest))
+	hexHashedCanonicalRequest := hex.EncodeToString(hashedCanonicalRequest[:])
+
+	region := "cn-north-1"
+	serviceName := "cv"
+	credentialScope := fmt.Sprintf("%s/%s/%s/request", shortDate, region, serviceName)
+	stringToSign := fmt.Sprintf("HMAC-SHA256\n%s\n%s\n%s",
+		xDate,
+		credentialScope,
+		hexHashedCanonicalRequest,
+	)
+
+	kDate := hmacSHA256([]byte(secretKey), []byte(shortDate))
+	kRegion := hmacSHA256(kDate, []byte(region))
+	kService := hmacSHA256(kRegion, []byte(serviceName))
+	kSigning := hmacSHA256(kService, []byte("request"))
+	signature := hex.EncodeToString(hmacSHA256(kSigning, []byte(stringToSign)))
+
+	authorization := fmt.Sprintf("HMAC-SHA256 Credential=%s/%s, SignedHeaders=%s, Signature=%s",
+		accessKey,
+		credentialScope,
+		signedHeaders,
+		signature,
+	)
+	req.Header.Set("Authorization", authorization)
+	return nil
+}
+
+func hmacSHA256(key []byte, data []byte) []byte {
+	h := hmac.New(sha256.New, key)
+	h.Write(data)
+	return h.Sum(nil)
+}
+
+func (a *TaskAdaptor) convertToRequestPayload(req *relaycommon.TaskSubmitReq) (*requestPayload, error) {
+	r := requestPayload{
+		ReqKey:      "jimeng_vgfm_i2v_l20",
+		Prompt:      req.Prompt,
+		AspectRatio: "16:9", // Default aspect ratio
+		Seed:        -1,     // Default to random
+	}
+
+	// Handle one-of image_urls or binary_data_base64
+	if req.Image != "" {
+		if strings.HasPrefix(req.Image, "http") {
+			r.ImageUrls = []string{req.Image}
+		} else {
+			r.BinaryDataBase64 = []string{req.Image}
+		}
+	}
+	metadata := req.Metadata
+	medaBytes, err := json.Marshal(metadata)
+	if err != nil {
+		return nil, errors.Wrap(err, "metadata marshal metadata failed")
+	}
+	err = json.Unmarshal(medaBytes, &r)
+	if err != nil {
+		return nil, errors.Wrap(err, "unmarshal metadata failed")
+	}
+	return &r, nil
+}
+
+func (a *TaskAdaptor) ParseTaskResult(respBody []byte) (*relaycommon.TaskInfo, error) {
+	resTask := responseTask{}
+	if err := json.Unmarshal(respBody, &resTask); err != nil {
+		return nil, errors.Wrap(err, "unmarshal task result failed")
+	}
+	taskResult := relaycommon.TaskInfo{}
+	if resTask.Code == 10000 {
+		taskResult.Code = 0
+	} else {
+		taskResult.Code = resTask.Code // todo uni code
+		taskResult.Reason = resTask.Message
+		taskResult.Status = model.TaskStatusFailure
+		taskResult.Progress = "100%"
+	}
+	switch resTask.Data.Status {
+	case "in_queue":
+		taskResult.Status = model.TaskStatusQueued
+		taskResult.Progress = "10%"
+	case "done":
+		taskResult.Status = model.TaskStatusSuccess
+		taskResult.Progress = "100%"
+	}
+	taskResult.Url = resTask.Data.VideoUrl
+	return &taskResult, nil
+}

relay/channel/task/kling/adaptor.go

@@ -2,11 +2,12 @@ package kling
 
 import (
 	"bytes"
-	"context"
 	"encoding/json"
 	"fmt"
+	"github.com/samber/lo"
 	"io"
 	"net/http"
+	"one-api/model"
 	"strings"
 	"time"
 
@@ -15,6 +16,7 @@ import (
 	"github.com/pkg/errors"
 
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
@@ -41,16 +43,27 @@ type requestPayload struct {
 	Mode        string  `json:"mode,omitempty"`
 	Duration    string  `json:"duration,omitempty"`
 	AspectRatio string  `json:"aspect_ratio,omitempty"`
-	Model       string  `json:"model,omitempty"`
 	ModelName   string  `json:"model_name,omitempty"`
 	CfgScale    float64 `json:"cfg_scale,omitempty"`
 }
 
 type responsePayload struct {
-	Code    int    `json:"code"`
-	Message string `json:"message"`
-	Data    struct {
-		TaskID string `json:"task_id"`
+	Code      int    `json:"code"`
+	Message   string `json:"message"`
+	RequestId string `json:"request_id"`
+	Data      struct {
+		TaskId        string `json:"task_id"`
+		TaskStatus    string `json:"task_status"`
+		TaskStatusMsg string `json:"task_status_msg"`
+		TaskResult    struct {
+			Videos []struct {
+				Id       string `json:"id"`
+				Url      string `json:"url"`
+				Duration string `json:"duration"`
+			} `json:"videos"`
+		} `json:"task_result"`
+		CreatedAt int64 `json:"created_at"`
+		UpdatedAt int64 `json:"updated_at"`
 	} `json:"data"`
 }
 
@@ -69,8 +82,8 @@ func (a *TaskAdaptor) Init(info *relaycommon.TaskRelayInfo) {
 	a.ChannelType = info.ChannelType
 	a.baseURL = info.BaseUrl
 
-	// apiKey format: "access_key,secret_key"
-	keyParts := strings.Split(info.ApiKey, ",")
+	// apiKey format: "access_key|secret_key"
+	keyParts := strings.Split(info.ApiKey, "|")
 	if len(keyParts) == 2 {
 		a.accessKey = strings.TrimSpace(keyParts[0])
 		a.secretKey = strings.TrimSpace(keyParts[1])
@@ -80,7 +93,7 @@ func (a *TaskAdaptor) Init(info *relaycommon.TaskRelayInfo) {
 // ValidateRequestAndSetAction parses body, validates fields and sets default action.
 func (a *TaskAdaptor) ValidateRequestAndSetAction(c *gin.Context, info *relaycommon.TaskRelayInfo) (taskErr *dto.TaskError) {
 	// Accept only POST /v1/video/generations as "generate" action.
-	action := "generate"
+	action := constant.TaskActionGenerate
 	info.Action = action
 
 	var req SubmitReq
@@ -94,13 +107,14 @@ func (a *TaskAdaptor) ValidateRequestAndSetAction(c *gin.Context, info *relaycom
 	}
 
 	// Store into context for later usage
-	c.Set("kling_request", req)
+	c.Set("task_request", req)
 	return nil
 }
 
 // BuildRequestURL constructs the upstream URL.
 func (a *TaskAdaptor) BuildRequestURL(info *relaycommon.TaskRelayInfo) (string, error) {
-	return fmt.Sprintf("%s/v1/videos/image2video", a.baseURL), nil
+	path := lo.Ternary(info.Action == constant.TaskActionGenerate, "/v1/videos/image2video", "/v1/videos/text2video")
+	return fmt.Sprintf("%s%s", a.baseURL, path), nil
 }
 
 // BuildRequestHeader sets required headers.
@@ -119,13 +133,16 @@ func (a *TaskAdaptor) BuildRequestHeader(c *gin.Context, req *http.Request, info
 
 // BuildRequestBody converts request into Kling specific format.
 func (a *TaskAdaptor) BuildRequestBody(c *gin.Context, info *relaycommon.TaskRelayInfo) (io.Reader, error) {
-	v, exists := c.Get("kling_request")
+	v, exists := c.Get("task_request")
 	if !exists {
 		return nil, fmt.Errorf("request not found in context")
 	}
 	req := v.(SubmitReq)
 
-	body := a.convertToRequestPayload(&req)
+	body, err := a.convertToRequestPayload(&req)
+	if err != nil {
+		return nil, err
+	}
 	data, err := json.Marshal(body)
 	if err != nil {
 		return nil, err
@@ -135,6 +152,9 @@ func (a *TaskAdaptor) BuildRequestBody(c *gin.Context, info *relaycommon.TaskRel
 
 // DoRequest delegates to common helper.
 func (a *TaskAdaptor) DoRequest(c *gin.Context, info *relaycommon.TaskRelayInfo, requestBody io.Reader) (*http.Response, error) {
+	if action := c.GetString("action"); action != "" {
+		info.Action = action
+	}
 	return channel.DoTaskApiRequest(a, c, info, requestBody)
 }
 
@@ -149,8 +169,8 @@ func (a *TaskAdaptor) DoResponse(c *gin.Context, resp *http.Response, info *rela
 	// Attempt Kling response parse first.
 	var kResp responsePayload
 	if err := json.Unmarshal(responseBody, &kResp); err == nil && kResp.Code == 0 {
-		c.JSON(http.StatusOK, gin.H{"task_id": kResp.Data.TaskID})
-		return kResp.Data.TaskID, responseBody, nil
+		c.JSON(http.StatusOK, gin.H{"task_id": kResp.Data.TaskId})
+		return kResp.Data.TaskId, responseBody, nil
 	}
 
 	// Fallback generic task response.
@@ -175,7 +195,12 @@ func (a *TaskAdaptor) FetchTask(baseUrl, key string, body map[string]any) (*http
 	if !ok {
 		return nil, fmt.Errorf("invalid task_id")
 	}
-	url := fmt.Sprintf("%s/v1/videos/image2video/%s", baseUrl, taskID)
+	action, ok := body["action"].(string)
+	if !ok {
+		return nil, fmt.Errorf("invalid action")
+	}
+	path := lo.Ternary(action == constant.TaskActionGenerate, "/v1/videos/image2video", "/v1/videos/text2video")
+	url := fmt.Sprintf("%s%s/%s", baseUrl, path, taskID)
 
 	req, err := http.NewRequest(http.MethodGet, url, nil)
 	if err != nil {
@@ -187,10 +212,6 @@ func (a *TaskAdaptor) FetchTask(baseUrl, key string, body map[string]any) (*http
 		token = key
 	}
 
-	ctx, cancel := context.WithTimeout(context.Background(), 15*time.Second)
-	defer cancel()
-
-	req = req.WithContext(ctx)
 	req.Header.Set("Accept", "application/json")
 	req.Header.Set("Authorization", "Bearer "+token)
 	req.Header.Set("User-Agent", "kling-sdk/1.0")
@@ -210,22 +231,29 @@ func (a *TaskAdaptor) GetChannelName() string {
 // helpers
 // ============================
 
-func (a *TaskAdaptor) convertToRequestPayload(req *SubmitReq) *requestPayload {
-	r := &requestPayload{
+func (a *TaskAdaptor) convertToRequestPayload(req *SubmitReq) (*requestPayload, error) {
+	r := requestPayload{
 		Prompt:      req.Prompt,
 		Image:       req.Image,
 		Mode:        defaultString(req.Mode, "std"),
 		Duration:    fmt.Sprintf("%d", defaultInt(req.Duration, 5)),
 		AspectRatio: a.getAspectRatio(req.Size),
-		Model:       req.Model,
 		ModelName:   req.Model,
 		CfgScale:    0.5,
 	}
-	if r.Model == "" {
-		r.Model = "kling-v1"
+	if r.ModelName == "" {
 		r.ModelName = "kling-v1"
 	}
-	return r
+	metadata := req.Metadata
+	medaBytes, err := json.Marshal(metadata)
+	if err != nil {
+		return nil, errors.Wrap(err, "metadata marshal metadata failed")
+	}
+	err = json.Unmarshal(medaBytes, &r)
+	if err != nil {
+		return nil, errors.Wrap(err, "unmarshal metadata failed")
+	}
+	return &r, nil
 }
 
 func (a *TaskAdaptor) getAspectRatio(size string) string {
@@ -264,7 +292,7 @@ func (a *TaskAdaptor) createJWTToken() (string, error) {
 }
 
 func (a *TaskAdaptor) createJWTTokenWithKey(apiKey string) (string, error) {
-	parts := strings.Split(apiKey, ",")
+	parts := strings.Split(apiKey, "|")
 	if len(parts) != 2 {
 		return "", fmt.Errorf("invalid API key format, expected 'access_key,secret_key'")
 	}
@@ -286,27 +314,33 @@ func (a *TaskAdaptor) createJWTTokenWithKeys(accessKey, secretKey string) (strin
 	return token.SignedString([]byte(secretKey))
 }
 
-// ParseResultUrl 提取视频任务结果的 url
-func (a *TaskAdaptor) ParseResultUrl(resp map[string]any) (string, error) {
-	data, ok := resp["data"].(map[string]any)
-	if !ok {
-		return "", fmt.Errorf("data field not found or invalid")
+func (a *TaskAdaptor) ParseTaskResult(respBody []byte) (*relaycommon.TaskInfo, error) {
+	resPayload := responsePayload{}
+	err := json.Unmarshal(respBody, &resPayload)
+	if err != nil {
+		return nil, errors.Wrap(err, "failed to unmarshal response body")
 	}
-	taskResult, ok := data["task_result"].(map[string]any)
-	if !ok {
-		return "", fmt.Errorf("task_result field not found or invalid")
+	taskInfo := &relaycommon.TaskInfo{}
+	taskInfo.Code = resPayload.Code
+	taskInfo.TaskID = resPayload.Data.TaskId
+	taskInfo.Reason = resPayload.Message
+	//任务状态，枚举值：submitted（已提交）、processing（处理中）、succeed（成功）、failed（失败）
+	status := resPayload.Data.TaskStatus
+	switch status {
+	case "submitted":
+		taskInfo.Status = model.TaskStatusSubmitted
+	case "processing":
+		taskInfo.Status = model.TaskStatusInProgress
+	case "succeed":
+		taskInfo.Status = model.TaskStatusSuccess
+	case "failed":
+		taskInfo.Status = model.TaskStatusFailure
+	default:
+		return nil, fmt.Errorf("unknown task status: %s", status)
 	}
-	videos, ok := taskResult["videos"].([]interface{})
-	if !ok || len(videos) == 0 {
-		return "", fmt.Errorf("videos field not found or empty")
+	if videos := resPayload.Data.TaskResult.Videos; len(videos) > 0 {
+		video := videos[0]
+		taskInfo.Url = video.Url
 	}
-	video, ok := videos[0].(map[string]interface{})
-	if !ok {
-		return "", fmt.Errorf("video item invalid")
-	}
-	url, ok := video["url"].(string)
-	if !ok || url == "" {
-		return "", fmt.Errorf("url field not found or invalid")
-	}
-	return url, nil
+	return taskInfo, nil
 }

relay/channel/task/suno/adaptor.go

@@ -22,8 +22,8 @@ type TaskAdaptor struct {
 	ChannelType int
 }
 
-func (a *TaskAdaptor) ParseResultUrl(resp map[string]any) (string, error) {
-	return "", nil // todo implement this method if needed
+func (a *TaskAdaptor) ParseTaskResult([]byte) (*relaycommon.TaskInfo, error) {
+	return nil, fmt.Errorf("not implement") // todo implement this method if needed
 }
 
 func (a *TaskAdaptor) Init(info *relaycommon.TaskRelayInfo) {

relay/channel/tencent/relay-tencent.go

@@ -124,10 +124,7 @@ func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 
 	helper.Done(c)
 
-	err := resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
-	}
+	common.CloseResponseBodyGracefully(resp)
 
 	return nil, responseText
 }
@@ -138,10 +135,7 @@ func tencentHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithSt
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &tencentSb)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/vertex/service_account.go

@@ -11,6 +11,7 @@ import (
 	"net/http"
 	"net/url"
 	relaycommon "one-api/relay/common"
+	"one-api/service"
 	"strings"
 
 	"fmt"
@@ -45,7 +46,7 @@ func getAccessToken(a *Adaptor, info *relaycommon.RelayInfo) (string, error) {
 	if err != nil {
 		return "", fmt.Errorf("failed to create signed JWT: %w", err)
 	}
-	newToken, err := exchangeJwtForAccessToken(signedJWT)
+	newToken, err := exchangeJwtForAccessToken(signedJWT, info)
 	if err != nil {
 		return "", fmt.Errorf("failed to exchange JWT for access token: %w", err)
 	}
@@ -96,14 +97,25 @@ func createSignedJWT(email, privateKeyPEM string) (string, error) {
 	return signedToken, nil
 }
 
-func exchangeJwtForAccessToken(signedJWT string) (string, error) {
+func exchangeJwtForAccessToken(signedJWT string, info *relaycommon.RelayInfo) (string, error) {
 
 	authURL := "https://www.googleapis.com/oauth2/v4/token"
 	data := url.Values{}
 	data.Set("grant_type", "urn:ietf:params:oauth:grant-type:jwt-bearer")
 	data.Set("assertion", signedJWT)
 
-	resp, err := http.PostForm(authURL, data)
+	var client *http.Client
+	var err error
+	if proxyURL, ok := info.ChannelSetting["proxy"]; ok {
+		client, err = service.NewProxyHttpClient(proxyURL.(string))
+		if err != nil {
+			return "", fmt.Errorf("new proxy http client failed: %w", err)
+		}
+	} else {
+		client = service.GetHttpClient()
+	}
+
+	resp, err := client.PostForm(authURL, data)
 	if err != nil {
 		return "", err
 	}

relay/channel/volcengine/adaptor.go

@@ -1,15 +1,19 @@
 package volcengine
 
 import (
+	"bytes"
 	"errors"
 	"fmt"
 	"io"
+	"mime/multipart"
 	"net/http"
+	"net/textproto"
 	"one-api/dto"
 	"one-api/relay/channel"
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"path/filepath"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -30,8 +34,146 @@ 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) {
-	//TODO implement me
-	return nil, errors.New("not implemented")
+	switch info.RelayMode {
+	case constant.RelayModeImagesEdits:
+
+		var requestBody bytes.Buffer
+		writer := multipart.NewWriter(&requestBody)
+
+		writer.WriteField("model", request.Model)
+		// 获取所有表单字段
+		formData := c.Request.PostForm
+		// 遍历表单字段并打印输出
+		for key, values := range formData {
+			if key == "model" {
+				continue
+			}
+			for _, value := range values {
+				writer.WriteField(key, value)
+			}
+		}
+
+		// Parse the multipart form to handle both single image and multiple images
+		if err := c.Request.ParseMultipartForm(32 << 20); err != nil { // 32MB max memory
+			return nil, errors.New("failed to parse multipart form")
+		}
+
+		if c.Request.MultipartForm != nil && c.Request.MultipartForm.File != nil {
+			// Check if "image" field exists in any form, including array notation
+			var imageFiles []*multipart.FileHeader
+			var exists bool
+
+			// First check for standard "image" field
+			if imageFiles, exists = c.Request.MultipartForm.File["image"]; !exists || len(imageFiles) == 0 {
+				// If not found, check for "image[]" field
+				if imageFiles, exists = c.Request.MultipartForm.File["image[]"]; !exists || len(imageFiles) == 0 {
+					// If still not found, iterate through all fields to find any that start with "image["
+					foundArrayImages := false
+					for fieldName, files := range c.Request.MultipartForm.File {
+						if strings.HasPrefix(fieldName, "image[") && len(files) > 0 {
+							foundArrayImages = true
+							for _, file := range files {
+								imageFiles = append(imageFiles, file)
+							}
+						}
+					}
+
+					// If no image fields found at all
+					if !foundArrayImages && (len(imageFiles) == 0) {
+						return nil, errors.New("image is required")
+					}
+				}
+			}
+
+			// Process all image files
+			for i, fileHeader := range imageFiles {
+				file, err := fileHeader.Open()
+				if err != nil {
+					return nil, fmt.Errorf("failed to open image file %d: %w", i, err)
+				}
+				defer file.Close()
+
+				// If multiple images, use image[] as the field name
+				fieldName := "image"
+				if len(imageFiles) > 1 {
+					fieldName = "image[]"
+				}
+
+				// Determine MIME type based on file extension
+				mimeType := detectImageMimeType(fileHeader.Filename)
+
+				// Create a form file with the appropriate content type
+				h := make(textproto.MIMEHeader)
+				h.Set("Content-Disposition", fmt.Sprintf(`form-data; name="%s"; filename="%s"`, fieldName, fileHeader.Filename))
+				h.Set("Content-Type", mimeType)
+
+				part, err := writer.CreatePart(h)
+				if err != nil {
+					return nil, fmt.Errorf("create form part failed for image %d: %w", i, err)
+				}
+
+				if _, err := io.Copy(part, file); err != nil {
+					return nil, fmt.Errorf("copy file failed for image %d: %w", i, err)
+				}
+			}
+
+			// Handle mask file if present
+			if maskFiles, exists := c.Request.MultipartForm.File["mask"]; exists && len(maskFiles) > 0 {
+				maskFile, err := maskFiles[0].Open()
+				if err != nil {
+					return nil, errors.New("failed to open mask file")
+				}
+				defer maskFile.Close()
+
+				// Determine MIME type for mask file
+				mimeType := detectImageMimeType(maskFiles[0].Filename)
+
+				// Create a form file with the appropriate content type
+				h := make(textproto.MIMEHeader)
+				h.Set("Content-Disposition", fmt.Sprintf(`form-data; name="mask"; filename="%s"`, maskFiles[0].Filename))
+				h.Set("Content-Type", mimeType)
+
+				maskPart, err := writer.CreatePart(h)
+				if err != nil {
+					return nil, errors.New("create form file failed for mask")
+				}
+
+				if _, err := io.Copy(maskPart, maskFile); err != nil {
+					return nil, errors.New("copy mask file failed")
+				}
+			}
+		} else {
+			return nil, errors.New("no multipart form data found")
+		}
+
+		// 关闭 multipart 编写器以设置分界线
+		writer.Close()
+		c.Request.Header.Set("Content-Type", writer.FormDataContentType())
+		return bytes.NewReader(requestBody.Bytes()), nil
+
+	default:
+		return request, nil
+	}
+}
+
+// detectImageMimeType determines the MIME type based on the file extension
+func detectImageMimeType(filename string) string {
+	ext := strings.ToLower(filepath.Ext(filename))
+	switch ext {
+	case ".jpg", ".jpeg":
+		return "image/jpeg"
+	case ".png":
+		return "image/png"
+	case ".webp":
+		return "image/webp"
+	default:
+		// Try to detect from extension if possible
+		if strings.HasPrefix(ext, ".jp") {
+			return "image/jpeg"
+		}
+		// Default to png as a fallback
+		return "image/png"
+	}
 }
 
 func (a *Adaptor) Init(info *relaycommon.RelayInfo) {
@@ -46,6 +188,8 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		return fmt.Sprintf("%s/api/v3/chat/completions", info.BaseUrl), nil
 	case constant.RelayModeEmbeddings:
 		return fmt.Sprintf("%s/api/v3/embeddings", info.BaseUrl), nil
+	case constant.RelayModeImagesGenerations:
+		return fmt.Sprintf("%s/api/v3/images/generations", info.BaseUrl), nil
 	default:
 	}
 	return "", fmt.Errorf("unsupported relay mode: %d", info.RelayMode)
@@ -91,6 +235,8 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 		}
 	case constant.RelayModeEmbeddings:
 		err, usage = openai.OpenaiHandler(c, resp, info)
+	case constant.RelayModeImagesGenerations, constant.RelayModeImagesEdits:
+		err, usage = openai.OpenaiHandlerWithUsage(c, resp, info)
 	}
 	return
 }

relay/channel/xai/text.go

@@ -1,9 +1,7 @@
 package xai
 
 import (
-	"bytes"
 	"encoding/json"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -13,6 +11,8 @@ import (
 	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
+
+	"github.com/gin-gonic/gin"
 )
 
 func streamResponseXAI2OpenAI(xAIResp *dto.ChatCompletionsStreamResponse, usage *dto.Usage) *dto.ChatCompletionsStreamResponse {
@@ -73,18 +73,16 @@ func xAIStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	}
 
 	helper.Done(c)
-	err := resp.Body.Close()
-	if err != nil {
-		//return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-		common.SysError("close_response_body_failed: " + err.Error())
-	}
+	common.CloseResponseBodyGracefully(resp)
 	return nil, usage
 }
 
 func xAIHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	defer common.CloseResponseBodyGracefully(resp)
+
 	responseBody, err := io.ReadAll(resp.Body)
-	var response *dto.TextResponse
-	err = common.DecodeJson(responseBody, &response)
+	var response *dto.SimpleResponse
+	err = common.UnmarshalJson(responseBody, &response)
 	if err != nil {
 		common.SysError("error unmarshalling stream response: " + err.Error())
 		return nil, nil
@@ -99,21 +97,7 @@ func xAIHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo
 		return nil, nil
 	}
 
-	// set new body
-	resp.Body = io.NopCloser(bytes.NewBuffer(encodeJson))
-
-	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
-	}
-	err = resp.Body.Close()
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
-	}
+	common.IOCopyBytesGracefully(c, resp, encodeJson)
 
 	return nil, &response.Usage
 }

relay/channel/xinference/dto.go

@@ -1,7 +1,7 @@
 package xinference
 
 type XinRerankResponseDocument struct {
-	Document       string  `json:"document,omitempty"`
+	Document       any     `json:"document,omitempty"`
 	Index          int     `json:"index"`
 	RelevanceScore float64 `json:"relevance_score"`
 }

relay/channel/zhipu/relay-zhipu.go

@@ -210,10 +210,7 @@ func zhipuStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWi
 			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
 }
 
@@ -223,10 +220,7 @@ func zhipuHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWithStat
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	err = json.Unmarshal(responseBody, &zhipuResponse)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/common/relay_info.go

@@ -65,8 +65,8 @@ type RelayInfo struct {
 	TokenId           int
 	TokenKey          string
 	UserId            int
-	Group             string
-	UserGroup         string
+	UsingGroup        string // 使用的分组
+	UserGroup         string // 用户所在分组
 	TokenUnlimited    bool
 	StartTime         time.Time
 	FirstResponseTime time.Time
@@ -113,17 +113,17 @@ type RelayInfo struct {
 
 // 定义支持流式选项的通道类型
 var streamSupportedChannels = map[int]bool{
-	common.ChannelTypeOpenAI:     true,
-	common.ChannelTypeAnthropic:  true,
-	common.ChannelTypeAws:        true,
-	common.ChannelTypeGemini:     true,
-	common.ChannelCloudflare:     true,
-	common.ChannelTypeAzure:      true,
-	common.ChannelTypeVolcEngine: true,
-	common.ChannelTypeOllama:     true,
-	common.ChannelTypeXai:        true,
-	common.ChannelTypeDeepSeek:   true,
-	common.ChannelTypeBaiduV2:    true,
+	constant.ChannelTypeOpenAI:     true,
+	constant.ChannelTypeAnthropic:  true,
+	constant.ChannelTypeAws:        true,
+	constant.ChannelTypeGemini:     true,
+	constant.ChannelCloudflare:     true,
+	constant.ChannelTypeAzure:      true,
+	constant.ChannelTypeVolcEngine: true,
+	constant.ChannelTypeOllama:     true,
+	constant.ChannelTypeXai:        true,
+	constant.ChannelTypeDeepSeek:   true,
+	constant.ChannelTypeBaiduV2:    true,
 }
 
 func GenRelayInfoWs(c *gin.Context, ws *websocket.Conn) *RelayInfo {
@@ -211,41 +211,40 @@ func GenRelayInfoImage(c *gin.Context) *RelayInfo {
 }
 
 func GenRelayInfo(c *gin.Context) *RelayInfo {
-	channelType := c.GetInt("channel_type")
-	channelId := c.GetInt("channel_id")
-	channelSetting := c.GetStringMap("channel_setting")
-	paramOverride := c.GetStringMap("param_override")
+	channelType := common.GetContextKeyInt(c, constant.ContextKeyChannelType)
+	channelId := common.GetContextKeyInt(c, constant.ContextKeyChannelId)
+	channelSetting := common.GetContextKeyStringMap(c, constant.ContextKeyChannelSetting)
+	paramOverride := common.GetContextKeyStringMap(c, constant.ContextKeyParamOverride)
 
-	tokenId := c.GetInt("token_id")
-	tokenKey := c.GetString("token_key")
-	userId := c.GetInt("id")
-	group := c.GetString("group")
-	tokenUnlimited := c.GetBool("token_unlimited_quota")
-	startTime := c.GetTime(constant.ContextKeyRequestStartTime)
+	tokenId := common.GetContextKeyInt(c, constant.ContextKeyTokenId)
+	tokenKey := common.GetContextKeyString(c, constant.ContextKeyTokenKey)
+	userId := common.GetContextKeyInt(c, constant.ContextKeyUserId)
+	tokenUnlimited := common.GetContextKeyBool(c, constant.ContextKeyTokenUnlimited)
+	startTime := common.GetContextKeyTime(c, constant.ContextKeyRequestStartTime)
 	// firstResponseTime = time.Now() - 1 second
 
-	apiType, _ := relayconstant.ChannelType2APIType(channelType)
+	apiType, _ := common.ChannelType2APIType(channelType)
 
 	info := &RelayInfo{
-		UserQuota:         c.GetInt(constant.ContextKeyUserQuota),
-		UserSetting:       c.GetStringMap(constant.ContextKeyUserSetting),
-		UserEmail:         c.GetString(constant.ContextKeyUserEmail),
+		UserQuota:         common.GetContextKeyInt(c, constant.ContextKeyUserQuota),
+		UserSetting:       common.GetContextKeyStringMap(c, constant.ContextKeyUserSetting),
+		UserEmail:         common.GetContextKeyString(c, constant.ContextKeyUserEmail),
 		isFirstResponse:   true,
 		RelayMode:         relayconstant.Path2RelayMode(c.Request.URL.Path),
-		BaseUrl:           c.GetString("base_url"),
+		BaseUrl:           common.GetContextKeyString(c, constant.ContextKeyBaseUrl),
 		RequestURLPath:    c.Request.URL.String(),
 		ChannelType:       channelType,
 		ChannelId:         channelId,
 		TokenId:           tokenId,
 		TokenKey:          tokenKey,
 		UserId:            userId,
-		Group:             group,
-		UserGroup:         c.GetString(constant.ContextKeyUserGroup),
+		UsingGroup:        common.GetContextKeyString(c, constant.ContextKeyUsingGroup),
+		UserGroup:         common.GetContextKeyString(c, constant.ContextKeyUserGroup),
 		TokenUnlimited:    tokenUnlimited,
 		StartTime:         startTime,
 		FirstResponseTime: startTime.Add(-time.Second),
-		OriginModelName:   c.GetString("original_model"),
-		UpstreamModelName: c.GetString("original_model"),
+		OriginModelName:   common.GetContextKeyString(c, constant.ContextKeyOriginalModel),
+		UpstreamModelName: common.GetContextKeyString(c, constant.ContextKeyOriginalModel),
 		//RecodeModelName:   c.GetString("original_model"),
 		IsModelMapped:     false,
 		ApiType:           apiType,
@@ -267,12 +266,12 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 		info.RequestURLPath = "/v1" + info.RequestURLPath
 	}
 	if info.BaseUrl == "" {
-		info.BaseUrl = common.ChannelBaseURLs[channelType]
+		info.BaseUrl = constant.ChannelBaseURLs[channelType]
 	}
-	if info.ChannelType == common.ChannelTypeAzure {
+	if info.ChannelType == constant.ChannelTypeAzure {
 		info.ApiVersion = GetAPIVersion(c)
 	}
-	if info.ChannelType == common.ChannelTypeVertexAi {
+	if info.ChannelType == constant.ChannelTypeVertexAi {
 		info.ApiVersion = c.GetString("region")
 	}
 	if streamSupportedChannels[info.ChannelType] {
@@ -314,3 +313,22 @@ func GenTaskRelayInfo(c *gin.Context) *TaskRelayInfo {
 	}
 	return info
 }
+
+type TaskSubmitReq struct {
+	Prompt   string                 `json:"prompt"`
+	Model    string                 `json:"model,omitempty"`
+	Mode     string                 `json:"mode,omitempty"`
+	Image    string                 `json:"image,omitempty"`
+	Size     string                 `json:"size,omitempty"`
+	Duration int                    `json:"duration,omitempty"`
+	Metadata map[string]interface{} `json:"metadata,omitempty"`
+}
+
+type TaskInfo struct {
+	Code     int    `json:"code"`
+	TaskID   string `json:"task_id"`
+	Status   string `json:"status"`
+	Reason   string `json:"reason,omitempty"`
+	Url      string `json:"url,omitempty"`
+	Progress string `json:"progress,omitempty"`
+}

relay/common/relay_utils.go

@@ -6,7 +6,7 @@ import (
 	_ "image/gif"
 	_ "image/jpeg"
 	_ "image/png"
-	"one-api/common"
+	"one-api/constant"
 	"strings"
 )
 
@@ -15,9 +15,9 @@ func GetFullRequestURL(baseURL string, requestURL string, channelType int) strin
 
 	if strings.HasPrefix(baseURL, "https://gateway.ai.cloudflare.com") {
 		switch channelType {
-		case common.ChannelTypeOpenAI:
+		case constant.ChannelTypeOpenAI:
 			fullRequestURL = fmt.Sprintf("%s%s", baseURL, strings.TrimPrefix(requestURL, "/v1"))
-		case common.ChannelTypeAzure:
+		case constant.ChannelTypeAzure:
 			fullRequestURL = fmt.Sprintf("%s%s", baseURL, strings.TrimPrefix(requestURL, "/openai/deployments"))
 		}
 	}

relay/common_handler/rerank.go

@@ -5,6 +5,7 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel/xinference"
 	relaycommon "one-api/relay/common"
@@ -16,17 +17,14 @@ func RerankHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Respo
 	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
-	}
+	common.CloseResponseBodyGracefully(resp)
 	if common.DebugEnabled {
 		println("reranker response body: ", string(responseBody))
 	}
 	var jinaResp dto.RerankResponse
-	if info.ChannelType == common.ChannelTypeXinference {
+	if info.ChannelType == constant.ChannelTypeXinference {
 		var xinRerankResponse xinference.XinRerankResponse
-		err = common.DecodeJson(responseBody, &xinRerankResponse)
+		err = common.UnmarshalJson(responseBody, &xinRerankResponse)
 		if err != nil {
 			return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 		}
@@ -38,10 +36,16 @@ func RerankHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Respo
 			}
 			if info.ReturnDocuments {
 				var document any
-				if result.Document == "" {
-					document = info.Documents[result.Index]
-				} else {
-					document = result.Document
+				if result.Document != nil {
+					if doc, ok := result.Document.(string); ok {
+						if doc == "" {
+							document = info.Documents[result.Index]
+						} else {
+							document = doc
+						}
+					} else {
+						document = result.Document
+					}
 				}
 				respResult.Document = document
 			}
@@ -55,7 +59,7 @@ func RerankHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Respo
 			},
 		}
 	} else {
-		err = common.DecodeJson(responseBody, &jinaResp)
+		err = common.UnmarshalJson(responseBody, &jinaResp)
 		if err != nil {
 			return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 		}

relay/constant/api_type.go

@@ -1,106 +0,0 @@
-package constant
-
-import (
-	"one-api/common"
-)
-
-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
-)
-
-func ChannelType2APIType(channelType int) (int, bool) {
-	apiType := -1
-	switch channelType {
-	case common.ChannelTypeOpenAI:
-		apiType = APITypeOpenAI
-	case common.ChannelTypeAnthropic:
-		apiType = APITypeAnthropic
-	case common.ChannelTypeBaidu:
-		apiType = APITypeBaidu
-	case common.ChannelTypePaLM:
-		apiType = APITypePaLM
-	case common.ChannelTypeZhipu:
-		apiType = APITypeZhipu
-	case common.ChannelTypeAli:
-		apiType = APITypeAli
-	case common.ChannelTypeXunfei:
-		apiType = APITypeXunfei
-	case common.ChannelTypeAIProxyLibrary:
-		apiType = APITypeAIProxyLibrary
-	case common.ChannelTypeTencent:
-		apiType = APITypeTencent
-	case common.ChannelTypeGemini:
-		apiType = APITypeGemini
-	case common.ChannelTypeZhipu_v4:
-		apiType = APITypeZhipuV4
-	case common.ChannelTypeOllama:
-		apiType = APITypeOllama
-	case common.ChannelTypePerplexity:
-		apiType = APITypePerplexity
-	case common.ChannelTypeAws:
-		apiType = APITypeAws
-	case common.ChannelTypeCohere:
-		apiType = APITypeCohere
-	case common.ChannelTypeDify:
-		apiType = APITypeDify
-	case common.ChannelTypeJina:
-		apiType = APITypeJina
-	case common.ChannelCloudflare:
-		apiType = APITypeCloudflare
-	case common.ChannelTypeSiliconFlow:
-		apiType = APITypeSiliconFlow
-	case common.ChannelTypeVertexAi:
-		apiType = APITypeVertexAi
-	case common.ChannelTypeMistral:
-		apiType = APITypeMistral
-	case common.ChannelTypeDeepSeek:
-		apiType = APITypeDeepSeek
-	case common.ChannelTypeMokaAI:
-		apiType = APITypeMokaAI
-	case common.ChannelTypeVolcEngine:
-		apiType = APITypeVolcEngine
-	case common.ChannelTypeBaiduV2:
-		apiType = APITypeBaiduV2
-	case common.ChannelTypeOpenRouter:
-		apiType = APITypeOpenRouter
-	case common.ChannelTypeXinference:
-		apiType = APITypeXinference
-	case common.ChannelTypeXai:
-		apiType = APITypeXai
-	case common.ChannelTypeCoze:
-		apiType = APITypeCoze
-	}
-	if apiType == -1 {
-		return APITypeOpenAI, false
-	}
-	return apiType, true
-}

relay/constant/relay_mode.go

@@ -41,6 +41,9 @@ const (
 	RelayModeKlingFetchByID
 	RelayModeKlingSubmit
 
+	RelayModeJimengFetchByID
+	RelayModeJimengSubmit
+
 	RelayModeRerank
 
 	RelayModeResponses
@@ -80,7 +83,7 @@ func Path2RelayMode(path string) int {
 		relayMode = RelayModeRerank
 	} else if strings.HasPrefix(path, "/v1/realtime") {
 		relayMode = RelayModeRealtime
-	} else if strings.HasPrefix(path, "/v1beta/models") {
+	} else if strings.HasPrefix(path, "/v1beta/models") || strings.HasPrefix(path, "/v1/models") {
 		relayMode = RelayModeGemini
 	}
 	return relayMode
@@ -146,3 +149,13 @@ func Path2RelayKling(method, path string) int {
 	}
 	return relayMode
 }
+
+func Path2RelayJimeng(method, path string) int {
+	relayMode := RelayModeUnknown
+	if method == http.MethodPost && strings.HasSuffix(path, "/video/generations") {
+		relayMode = RelayModeJimengSubmit
+	} else if method == http.MethodGet && strings.Contains(path, "/video/generations/") {
+		relayMode = RelayModeJimengFetchByID
+	}
+	return relayMode
+}

relay/gemini_handler.go

@@ -13,6 +13,7 @@ import (
 	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
+	"one-api/setting/model_setting"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -76,6 +77,33 @@ func getGeminiInputTokens(req *gemini.GeminiChatRequest, info *relaycommon.Relay
 	return inputTokens
 }
 
+func isNoThinkingRequest(req *gemini.GeminiChatRequest) bool {
+	if req.GenerationConfig.ThinkingConfig != nil && req.GenerationConfig.ThinkingConfig.ThinkingBudget != nil {
+		return *req.GenerationConfig.ThinkingConfig.ThinkingBudget <= 0
+	}
+	return false
+}
+
+func trimModelThinking(modelName string) string {
+	// 去除模型名称中的 -nothinking 后缀
+	if strings.HasSuffix(modelName, "-nothinking") {
+		return strings.TrimSuffix(modelName, "-nothinking")
+	}
+	// 去除模型名称中的 -thinking 后缀
+	if strings.HasSuffix(modelName, "-thinking") {
+		return strings.TrimSuffix(modelName, "-thinking")
+	}
+
+	// 去除模型名称中的 -thinking-number
+	if strings.Contains(modelName, "-thinking-") {
+		parts := strings.Split(modelName, "-thinking-")
+		if len(parts) > 1 {
+			return parts[0] + "-thinking"
+		}
+	}
+	return modelName
+}
+
 func GeminiHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	req, err := getAndValidateGeminiRequest(c)
 	if err != nil {
@@ -107,12 +135,27 @@ func GeminiHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		relayInfo.SetPromptTokens(promptTokens)
 	} else {
 		promptTokens := getGeminiInputTokens(req, relayInfo)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "count_input_tokens_error", http.StatusBadRequest)
-		}
 		c.Set("prompt_tokens", promptTokens)
 	}
 
+	if model_setting.GetGeminiSettings().ThinkingAdapterEnabled {
+		if isNoThinkingRequest(req) {
+			// check is thinking
+			if !strings.Contains(relayInfo.OriginModelName, "-nothinking") {
+				// try to get no thinking model price
+				noThinkingModelName := relayInfo.OriginModelName + "-nothinking"
+				containPrice := helper.ContainPriceOrRatio(noThinkingModelName)
+				if containPrice {
+					relayInfo.OriginModelName = noThinkingModelName
+					relayInfo.UpstreamModelName = noThinkingModelName
+				}
+			}
+		}
+		if req.GenerationConfig.ThinkingConfig == nil {
+			gemini.ThinkingAdaptor(req, relayInfo)
+		}
+	}
+
 	priceData, err := helper.ModelPriceHelper(c, relayInfo, relayInfo.PromptTokens, int(req.GenerationConfig.MaxOutputTokens))
 	if err != nil {
 		return service.OpenAIErrorWrapperLocal(err, "model_price_error", http.StatusInternalServerError)

relay/helper/price.go

@@ -13,6 +13,7 @@ import (
 type GroupRatioInfo struct {
 	GroupRatio        float64
 	GroupSpecialRatio float64
+	HasSpecialRatio   bool
 }
 
 type PriceData struct {
@@ -31,7 +32,7 @@ func (p PriceData) ToSetting() string {
 	return fmt.Sprintf("ModelPrice: %f, ModelRatio: %f, CompletionRatio: %f, CacheRatio: %f, GroupRatio: %f, UsePrice: %t, CacheCreationRatio: %f, ShouldPreConsumedQuota: %d, ImageRatio: %f", p.ModelPrice, p.ModelRatio, p.CompletionRatio, p.CacheRatio, p.GroupRatioInfo.GroupRatio, p.UsePrice, p.CacheCreationRatio, p.ShouldPreConsumedQuota, p.ImageRatio)
 }
 
-// HandleGroupRatio checks for "auto_group" in the context and updates the group ratio and relayInfo.Group if present
+// HandleGroupRatio checks for "auto_group" in the context and updates the group ratio and relayInfo.UsingGroup if present
 func HandleGroupRatio(ctx *gin.Context, relayInfo *relaycommon.RelayInfo) GroupRatioInfo {
 	groupRatioInfo := GroupRatioInfo{
 		GroupRatio:        1.0, // default ratio
@@ -44,18 +45,19 @@ func HandleGroupRatio(ctx *gin.Context, relayInfo *relaycommon.RelayInfo) GroupR
 		if common.DebugEnabled {
 			println(fmt.Sprintf("final group: %s", autoGroup))
 		}
-		relayInfo.Group = autoGroup.(string)
+		relayInfo.UsingGroup = autoGroup.(string)
 	}
 
 	// check user group special ratio
-	userGroupRatio, ok := ratio_setting.GetGroupGroupRatio(relayInfo.UserGroup, relayInfo.Group)
+	userGroupRatio, ok := ratio_setting.GetGroupGroupRatio(relayInfo.UserGroup, relayInfo.UsingGroup)
 	if ok {
 		// user group special ratio
 		groupRatioInfo.GroupSpecialRatio = userGroupRatio
 		groupRatioInfo.GroupRatio = userGroupRatio
+		groupRatioInfo.HasSpecialRatio = true
 	} else {
 		// normal group ratio
-		groupRatioInfo.GroupRatio = ratio_setting.GetGroupRatio(relayInfo.Group)
+		groupRatioInfo.GroupRatio = ratio_setting.GetGroupRatio(relayInfo.UsingGroup)
 	}
 
 	return groupRatioInfo
@@ -120,6 +122,35 @@ func ModelPriceHelper(c *gin.Context, info *relaycommon.RelayInfo, promptTokens
 	return priceData, nil
 }
 
+type PerCallPriceData struct {
+	ModelPrice     float64
+	Quota          int
+	GroupRatioInfo GroupRatioInfo
+}
+
+// ModelPriceHelperPerCall 按次计费的 PriceHelper (MJ、Task)
+func ModelPriceHelperPerCall(c *gin.Context, info *relaycommon.RelayInfo) PerCallPriceData {
+	groupRatioInfo := HandleGroupRatio(c, info)
+
+	modelPrice, success := ratio_setting.GetModelPrice(info.OriginModelName, true)
+	// 如果没有配置价格，则使用默认价格
+	if !success {
+		defaultPrice, ok := ratio_setting.GetDefaultModelRatioMap()[info.OriginModelName]
+		if !ok {
+			modelPrice = 0.1
+		} else {
+			modelPrice = defaultPrice
+		}
+	}
+	quota := int(modelPrice * common.QuotaPerUnit * groupRatioInfo.GroupRatio)
+	priceData := PerCallPriceData{
+		ModelPrice:     modelPrice,
+		Quota:          quota,
+		GroupRatioInfo: groupRatioInfo,
+	}
+	return priceData
+}
+
 func ContainPriceOrRatio(modelName string) bool {
 	_, ok := ratio_setting.GetModelPrice(modelName, false)
 	if ok {

relay/helper/stream_scanner.go

@@ -20,8 +20,8 @@ import (
 )
 
 const (
-	InitialScannerBufferSize = 64 << 10  // 64KB (64*1024)
-	MaxScannerBufferSize     = 10 << 20  // 10MB (10*1024*1024)
+	InitialScannerBufferSize = 64 << 10 // 64KB (64*1024)
+	MaxScannerBufferSize     = 10 << 20 // 10MB (10*1024*1024)
 	DefaultPingInterval      = 10 * time.Second
 )
 
@@ -49,7 +49,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 		scanner    = bufio.NewScanner(resp.Body)
 		ticker     = time.NewTicker(streamingTimeout)
 		pingTicker *time.Ticker
-		writeMutex sync.Mutex // Mutex to protect concurrent writes
+		writeMutex sync.Mutex     // Mutex to protect concurrent writes
 		wg         sync.WaitGroup // 用于等待所有 goroutine 退出
 	)
 
@@ -64,32 +64,39 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 		pingTicker = time.NewTicker(pingInterval)
 	}
 
+	if common.DebugEnabled {
+		// print timeout and ping interval for debugging
+		println("relay timeout seconds:", common.RelayTimeout)
+		println("streaming timeout seconds:", int64(streamingTimeout.Seconds()))
+		println("ping interval seconds:", int64(pingInterval.Seconds()))
+	}
+
 	// 改进资源清理，确保所有 goroutine 正确退出
 	defer func() {
 		// 通知所有 goroutine 停止
 		common.SafeSendBool(stopChan, true)
-		
+
 		ticker.Stop()
 		if pingTicker != nil {
 			pingTicker.Stop()
 		}
-		
+
 		// 等待所有 goroutine 退出，最多等待5秒
 		done := make(chan struct{})
 		go func() {
 			wg.Wait()
 			close(done)
 		}()
-		
+
 		select {
 		case <-done:
 		case <-time.After(5 * time.Second):
 			common.LogError(c, "timeout waiting for goroutines to exit")
 		}
-		
+
 		close(stopChan)
 	}()
-	
+
 	scanner.Buffer(make([]byte, InitialScannerBufferSize), MaxScannerBufferSize)
 	scanner.Split(bufio.ScanLines)
 	SetEventStreamHeaders(c)
@@ -113,12 +120,12 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 					println("ping goroutine exited")
 				}
 			}()
-			
+
 			// 添加超时保护，防止 goroutine 无限运行
 			maxPingDuration := 30 * time.Minute // 最大 ping 持续时间
 			pingTimeout := time.NewTimer(maxPingDuration)
 			defer pingTimeout.Stop()
-			
+
 			for {
 				select {
 				case <-pingTicker.C:
@@ -129,7 +136,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 						defer writeMutex.Unlock()
 						done <- PingData(c)
 					}()
-					
+
 					select {
 					case err := <-done:
 						if err != nil {
@@ -175,7 +182,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 				println("scanner goroutine exited")
 			}
 		}()
-		
+
 		for scanner.Scan() {
 			// 检查是否需要停止
 			select {
@@ -187,7 +194,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 				return
 			default:
 			}
-			
+
 			ticker.Reset(streamingTimeout)
 			data := scanner.Text()
 			if common.DebugEnabled {
@@ -205,7 +212,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 			data = strings.TrimSuffix(data, "\r")
 			if !strings.HasPrefix(data, "[DONE]") {
 				info.SetFirstResponseTime()
-				
+
 				// 使用超时机制防止写操作阻塞
 				done := make(chan bool, 1)
 				go func() {
@@ -213,7 +220,7 @@ func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon
 					defer writeMutex.Unlock()
 					done <- dataHandler(data)
 				}()
-				
+
 				select {
 				case success := <-done:
 					if !success {

relay/image_handler.go

@@ -8,6 +8,7 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/model"
 	relaycommon "one-api/relay/common"
@@ -44,6 +45,11 @@ func getAndValidImageRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.
 		if imageRequest.N == 0 {
 			imageRequest.N = 1
 		}
+
+		if info.ApiType == constant.APITypeVolcEngine {
+			watermark := formData.Has("watermark")
+			imageRequest.Watermark = &watermark
+		}
 	default:
 		err := common.UnmarshalBodyReusable(c, imageRequest)
 		if err != nil {

relay/relay-mj.go

@@ -13,9 +13,9 @@ import (
 	"one-api/model"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
-	"one-api/setting/ratio_setting"
 	"strconv"
 	"strings"
 	"time"
@@ -174,18 +174,9 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 		return service.MidjourneyErrorWrapper(constant.MjRequestError, "sour_base64_and_target_base64_is_required")
 	}
 	modelName := service.CoverActionToModelName(constant.MjActionSwapFace)
-	modelPrice, success := ratio_setting.GetModelPrice(modelName, true)
-	// 如果没有配置价格，则使用默认价格
-	if !success {
-		defaultPrice, ok := ratio_setting.GetDefaultModelRatioMap()[modelName]
-		if !ok {
-			modelPrice = 0.1
-		} else {
-			modelPrice = defaultPrice
-		}
-	}
-	groupRatio := ratio_setting.GetGroupRatio(group)
-	ratio := modelPrice * groupRatio
+
+	priceData := helper.ModelPriceHelperPerCall(c, relayInfo)
+
 	userQuota, err := model.GetUserQuota(userId, false)
 	if err != nil {
 		return &dto.MidjourneyResponse{
@@ -193,9 +184,8 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 			Description: err.Error(),
 		}
 	}
-	quota := int(ratio * common.QuotaPerUnit)
 
-	if userQuota-quota < 0 {
+	if userQuota-priceData.Quota < 0 {
 		return &dto.MidjourneyResponse{
 			Code:        4,
 			Description: "quota_not_enough",
@@ -210,26 +200,18 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 	}
 	defer func() {
 		if mjResp.StatusCode == 200 && mjResp.Response.Code == 1 {
-			err := service.PostConsumeQuota(relayInfo, quota, 0, true)
+			err := service.PostConsumeQuota(relayInfo, priceData.Quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
-			//err = model.CacheUpdateUserQuota(userId)
-			if err != nil {
-				common.SysError("error update user quota cache: " + err.Error())
-			}
-			if quota != 0 {
-				tokenName := c.GetString("token_name")
-				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, groupRatio, constant.MjActionSwapFace)
-				other := make(map[string]interface{})
-				other["model_price"] = modelPrice
-				other["group_ratio"] = groupRatio
-				model.RecordConsumeLog(c, userId, channelId, 0, 0, modelName, tokenName,
-					quota, logContent, tokenId, userQuota, 0, false, group, other)
-				model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
-				channelId := c.GetInt("channel_id")
-				model.UpdateChannelUsedQuota(channelId, quota)
-			}
+
+			tokenName := c.GetString("token_name")
+			logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", priceData.ModelPrice, priceData.GroupRatioInfo.GroupRatio, constant.MjActionSwapFace)
+			other := service.GenerateMjOtherInfo(priceData)
+			model.RecordConsumeLog(c, userId, channelId, 0, 0, modelName, tokenName,
+				priceData.Quota, logContent, tokenId, userQuota, 0, false, group, other)
+			model.UpdateUserUsedQuotaAndRequestCount(userId, priceData.Quota)
+			model.UpdateChannelUsedQuota(channelId, priceData.Quota)
 		}
 	}()
 	midjResponse := &mjResp.Response
@@ -250,7 +232,7 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 		Progress:    "0%",
 		FailReason:  "",
 		ChannelId:   c.GetInt("channel_id"),
-		Quota:       quota,
+		Quota:       priceData.Quota,
 	}
 	err = midjourneyTask.Insert()
 	if err != nil {
@@ -297,10 +279,7 @@ func RelayMidjourneyTaskImageSeed(c *gin.Context) *dto.MidjourneyResponse {
 	if err != nil {
 		return service.MidjourneyErrorWrapper(constant.MjRequestError, "unmarshal_response_body_failed")
 	}
-	_, err = io.Copy(c.Writer, bytes.NewBuffer(respBody))
-	if err != nil {
-		return service.MidjourneyErrorWrapper(constant.MjRequestError, "copy_response_body_failed")
-	}
+	common.IOCopyBytesGracefully(c, nil, respBody)
 	return nil
 }
 
@@ -480,18 +459,9 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 	fullRequestURL := fmt.Sprintf("%s%s", baseURL, requestURL)
 
 	modelName := service.CoverActionToModelName(midjRequest.Action)
-	modelPrice, success := ratio_setting.GetModelPrice(modelName, true)
-	// 如果没有配置价格，则使用默认价格
-	if !success {
-		defaultPrice, ok := ratio_setting.GetDefaultModelRatioMap()[modelName]
-		if !ok {
-			modelPrice = 0.1
-		} else {
-			modelPrice = defaultPrice
-		}
-	}
-	groupRatio := ratio_setting.GetGroupRatio(group)
-	ratio := modelPrice * groupRatio
+
+	priceData := helper.ModelPriceHelperPerCall(c, relayInfo)
+
 	userQuota, err := model.GetUserQuota(userId, false)
 	if err != nil {
 		return &dto.MidjourneyResponse{
@@ -499,9 +469,8 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 			Description: err.Error(),
 		}
 	}
-	quota := int(ratio * common.QuotaPerUnit)
 
-	if consumeQuota && userQuota-quota < 0 {
+	if consumeQuota && userQuota-priceData.Quota < 0 {
 		return &dto.MidjourneyResponse{
 			Code:        4,
 			Description: "quota_not_enough",
@@ -516,22 +485,17 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 
 	defer func() {
 		if consumeQuota && midjResponseWithStatus.StatusCode == 200 {
-			err := service.PostConsumeQuota(relayInfo, quota, 0, true)
+			err := service.PostConsumeQuota(relayInfo, priceData.Quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
-			if quota != 0 {
-				tokenName := c.GetString("token_name")
-				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s，ID %s", modelPrice, groupRatio, midjRequest.Action, midjResponse.Result)
-				other := make(map[string]interface{})
-				other["model_price"] = modelPrice
-				other["group_ratio"] = groupRatio
-				model.RecordConsumeLog(c, userId, channelId, 0, 0, modelName, tokenName,
-					quota, logContent, tokenId, userQuota, 0, false, group, other)
-				model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
-				channelId := c.GetInt("channel_id")
-				model.UpdateChannelUsedQuota(channelId, quota)
-			}
+			tokenName := c.GetString("token_name")
+			logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s，ID %s", priceData.ModelPrice, priceData.GroupRatioInfo.GroupRatio, midjRequest.Action, midjResponse.Result)
+			other := service.GenerateMjOtherInfo(priceData)
+			model.RecordConsumeLog(c, userId, channelId, 0, 0, modelName, tokenName,
+				priceData.Quota, logContent, tokenId, userQuota, 0, false, group, other)
+			model.UpdateUserUsedQuotaAndRequestCount(userId, priceData.Quota)
+			model.UpdateChannelUsedQuota(channelId, priceData.Quota)
 		}
 	}()
 
@@ -559,7 +523,7 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 		Progress:    "0%",
 		FailReason:  "",
 		ChannelId:   c.GetInt("channel_id"),
-		Quota:       quota,
+		Quota:       priceData.Quota,
 	}
 	if midjResponse.Code == 3 {
 		//无实例账号自动禁用渠道（No available account instance）

relay/relay-text.go

@@ -541,5 +541,5 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 		other["audio_input_price"] = audioInputPrice
 	}
 	model.RecordConsumeLog(ctx, relayInfo.UserId, relayInfo.ChannelId, promptTokens, completionTokens, logModel,
-		tokenName, quota, logContent, relayInfo.TokenId, userQuota, int(useTimeSeconds), relayInfo.IsStream, relayInfo.Group, other)
+		tokenName, quota, logContent, relayInfo.TokenId, userQuota, int(useTimeSeconds), relayInfo.IsStream, relayInfo.UsingGroup, other)
 }

relay/relay_adaptor.go

@@ -1,6 +1,7 @@
 package relay
 
 import (
+	"one-api/constant"
 	commonconstant "one-api/constant"
 	"one-api/relay/channel"
 	"one-api/relay/channel/ali"
@@ -22,6 +23,7 @@ import (
 	"one-api/relay/channel/palm"
 	"one-api/relay/channel/perplexity"
 	"one-api/relay/channel/siliconflow"
+	"one-api/relay/channel/task/jimeng"
 	"one-api/relay/channel/task/kling"
 	"one-api/relay/channel/task/suno"
 	"one-api/relay/channel/tencent"
@@ -31,7 +33,6 @@ import (
 	"one-api/relay/channel/xunfei"
 	"one-api/relay/channel/zhipu"
 	"one-api/relay/channel/zhipu_4v"
-	"one-api/relay/constant"
 )
 
 func GetAdaptor(apiType int) channel.Adaptor {
@@ -104,6 +105,8 @@ func GetTaskAdaptor(platform commonconstant.TaskPlatform) channel.TaskAdaptor {
 		return &suno.TaskAdaptor{}
 	case commonconstant.TaskPlatformKling:
 		return &kling.TaskAdaptor{}
+	case commonconstant.TaskPlatformJimeng:
+		return &jimeng.TaskAdaptor{}
 	}
 	return nil
 }

relay/relay_task.go

@@ -5,7 +5,6 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/common"
@@ -16,6 +15,8 @@ import (
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
 	"one-api/setting/ratio_setting"
+
+	"github.com/gin-gonic/gin"
 )
 
 /*
@@ -51,8 +52,14 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 	}
 
 	// 预扣
-	groupRatio := ratio_setting.GetGroupRatio(relayInfo.Group)
-	ratio := modelPrice * groupRatio
+	groupRatio := ratio_setting.GetGroupRatio(relayInfo.UsingGroup)
+	var ratio float64
+	userGroupRatio, hasUserGroupRatio := ratio_setting.GetGroupGroupRatio(relayInfo.UserGroup, relayInfo.UsingGroup)
+	if hasUserGroupRatio {
+		ratio = modelPrice * userGroupRatio
+	} else {
+		ratio = modelPrice * groupRatio
+	}
 	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		taskErr = service.TaskErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
@@ -121,12 +128,19 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 			}
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
-				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, groupRatio, relayInfo.Action)
+				gRatio := groupRatio
+				if hasUserGroupRatio {
+					gRatio = userGroupRatio
+				}
+				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, gRatio, relayInfo.Action)
 				other := make(map[string]interface{})
 				other["model_price"] = modelPrice
 				other["group_ratio"] = groupRatio
+				if hasUserGroupRatio {
+					other["user_group_ratio"] = userGroupRatio
+				}
 				model.RecordConsumeLog(c, relayInfo.UserId, relayInfo.ChannelId, 0, 0,
-					modelName, tokenName, quota, logContent, relayInfo.TokenId, userQuota, 0, false, relayInfo.Group, other)
+					modelName, tokenName, quota, logContent, relayInfo.TokenId, userQuota, 0, false, relayInfo.UsingGroup, other)
 				model.UpdateUserUsedQuotaAndRequestCount(relayInfo.UserId, quota)
 				model.UpdateChannelUsedQuota(relayInfo.ChannelId, quota)
 			}
@@ -231,7 +245,7 @@ func sunoFetchByIDRespBodyBuilder(c *gin.Context) (respBody []byte, taskResp *dt
 }
 
 func videoFetchByIDRespBodyBuilder(c *gin.Context) (respBody []byte, taskResp *dto.TaskError) {
-	taskId := c.Param("id")
+	taskId := c.Param("task_id")
 	userId := c.GetInt("id")
 
 	originTask, exist, err := model.GetByTaskId(userId, taskId)

router/api-router.go

@@ -125,6 +125,7 @@ func SetApiRouter(router *gin.Engine) {
 			tokenRoute.POST("/", controller.AddToken)
 			tokenRoute.PUT("/", controller.UpdateToken)
 			tokenRoute.DELETE("/:id", controller.DeleteToken)
+			tokenRoute.POST("/batch", controller.DeleteTokenBatch)
 		}
 		redemptionRoute := apiRouter.Group("/redemption")
 		redemptionRoute.Use(middleware.AdminAuth())

router/relay-router.go

@@ -63,6 +63,7 @@ func SetRelayRouter(router *gin.Engine) {
 		httpRouter.DELETE("/models/:model", controller.RelayNotImplemented)
 		httpRouter.POST("/moderations", controller.Relay)
 		httpRouter.POST("/rerank", controller.Relay)
+		httpRouter.POST("/models/*path", controller.Relay)
 	}
 
 	relayMjRouter := router.Group("/mj")

router/video-router.go

@@ -14,4 +14,11 @@ func SetVideoRouter(router *gin.Engine) {
 		videoV1Router.POST("/video/generations", controller.RelayTask)
 		videoV1Router.GET("/video/generations/:task_id", controller.RelayTask)
 	}
+
+	klingV1Router := router.Group("/kling/v1")
+	klingV1Router.Use(middleware.KlingRequestConvert(), middleware.TokenAuth(), middleware.Distribute())
+	{
+		klingV1Router.POST("/videos/text2video", controller.RelayTask)
+		klingV1Router.POST("/videos/image2video", controller.RelayTask)
+	}
 }

service/channel.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/model"
 	"one-api/setting/operation_setting"
@@ -48,7 +49,7 @@ func ShouldDisableChannel(channelType int, err *dto.OpenAIErrorWithStatusCode) b
 	}
 	if err.StatusCode == http.StatusForbidden {
 		switch channelType {
-		case common.ChannelTypeGemini:
+		case constant.ChannelTypeGemini:
 			return true
 		}
 	}

service/convert.go

@@ -4,6 +4,7 @@ import (
 	"encoding/json"
 	"fmt"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel/openrouter"
 	relaycommon "one-api/relay/common"
@@ -19,7 +20,7 @@ func ClaudeToOpenAIRequest(claudeRequest dto.ClaudeRequest, info *relaycommon.Re
 		Stream:      claudeRequest.Stream,
 	}
 
-	isOpenRouter := info.ChannelType == common.ChannelTypeOpenRouter
+	isOpenRouter := info.ChannelType == constant.ChannelTypeOpenRouter
 
 	if claudeRequest.Thinking != nil && claudeRequest.Thinking.Type == "enabled" {
 		if isOpenRouter {
@@ -276,12 +277,15 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 		}
 		if info.Done {
 			claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
-			if info.ClaudeConvertInfo.Usage != nil {
+			oaiUsage := info.ClaudeConvertInfo.Usage
+			if oaiUsage != nil {
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
 					Type: "message_delta",
 					Usage: &dto.ClaudeUsage{
-						InputTokens:  info.ClaudeConvertInfo.Usage.PromptTokens,
-						OutputTokens: info.ClaudeConvertInfo.Usage.CompletionTokens,
+						InputTokens:              oaiUsage.PromptTokens,
+						OutputTokens:             oaiUsage.CompletionTokens,
+						CacheCreationInputTokens: oaiUsage.PromptTokensDetails.CachedCreationTokens,
+						CacheReadInputTokens:     oaiUsage.PromptTokensDetails.CachedTokens,
 					},
 					Delta: &dto.ClaudeMediaMessage{
 						StopReason: common.GetPointer[string](stopReasonOpenAI2Claude(info.FinishReason)),