Merge pull request #820 from Calcium-Ion/cache

feat: Implement cache token ratio for more precise token pricing

README.en.md

@@ -68,7 +68,7 @@
 
 ## Model Support
 This version additionally supports:
-1. Third-party model **gps** (gpt-4-gizmo-*)
+1. Third-party model **gpts** (gpt-4-gizmo-*)
 2. [Midjourney-Proxy(Plus)](https://github.com/novicezk/midjourney-proxy) interface, [Integration Guide](Midjourney.md)
 3. Custom channels with full API URL support
 4. [Suno API](https://github.com/Suno-API/Suno-API) interface, [Integration Guide](Suno.md)
@@ -162,7 +162,7 @@ docker run --rm -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtow
 
 ## Channel Retry
 Channel retry is implemented, configurable in `Settings->Operation Settings->General Settings`. **Cache recommended**.  
-First retry uses same priority, second retry uses next priority, and so on.
+If retry is enabled, the system will automatically use the next priority channel for the same request after a failed request.
 
 ### Cache Configuration
 1. `REDIS_CONN_STRING`: Use Redis as cache

README.md

@@ -77,7 +77,7 @@
 
 ## 模型支持
 此版本额外支持以下模型：
-1. 第三方模型 **gps** （gpt-4-gizmo-*）
+1. 第三方模型 **gpts** （gpt-4-gizmo-*）
 2. [Midjourney-Proxy(Plus)](https://github.com/novicezk/midjourney-proxy)接口，[对接文档](Midjourney.md)
 3. 自定义渠道，支持填入完整调用地址
 4. [Suno API](https://github.com/Suno-API/Suno-API) 接口，[对接文档](Suno.md)
@@ -177,7 +177,7 @@ docker run --rm -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtow
 
 ## 渠道重试
 渠道重试功能已经实现，可以在`设置->运营设置->通用设置`设置重试次数，**建议开启缓存**功能。  
-如果开启了重试功能，第一次重试使用同优先级，第二次重试使用下一个优先级，以此类推。
+如果开启了重试功能，重试使用下一个优先级，以此类推。
 ### 缓存设置方法
 1. `REDIS_CONN_STRING`：设置之后将使用 Redis 作为缓存使用。
     + 例子：`REDIS_CONN_STRING=redis://default:redispw@localhost:49153`
@@ -220,8 +220,8 @@ docker run --rm -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtow
 - [neko-api-key-tool](https://github.com/Calcium-Ion/neko-api-key-tool)：用key查询使用额度
 
 其他基于New API的项目：
-- [new-api-horizon](https://github.com/Calcium-Ion/new-api-horizon)：New API高性能优化版，并支持Claude格式
-- [VoAPI](https://github.com/VoAPI/VoAPI)：基于New API的闭源项目
+- [new-api-horizon](https://github.com/Calcium-Ion/new-api-horizon)：New API高性能优化版，专注于高并发优化，并支持Claude格式
+- [VoAPI](https://github.com/VoAPI/VoAPI)：基于New API的前端美化版本，闭源免费
 
 ## 🌟 Star History
 

common/gopool.go

@@ -0,0 +1,24 @@
+package common
+
+import (
+	"context"
+	"fmt"
+	"github.com/bytedance/gopkg/util/gopool"
+	"math"
+)
+
+var relayGoPool gopool.Pool
+
+func init() {
+	relayGoPool = gopool.NewPool("gopool.RelayPool", math.MaxInt32, gopool.NewConfig())
+	relayGoPool.SetPanicHandler(func(ctx context.Context, i interface{}) {
+		if stopChan, ok := ctx.Value("stop_chan").(chan bool); ok {
+			SafeSendBool(stopChan, true)
+		}
+		SysError(fmt.Sprintf("panic in gopool.RelayPool: %v", i))
+	})
+}
+
+func RelayCtxGo(ctx context.Context, f func()) {
+	relayGoPool.CtxGo(ctx, f)
+}

controller/channel-test.go

@@ -17,6 +17,7 @@ import (
 	"one-api/relay"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strconv"
 	"strings"
@@ -72,18 +73,6 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 		}
 	}
 
-	modelMapping := *channel.ModelMapping
-	if modelMapping != "" && modelMapping != "{}" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return err, service.OpenAIErrorWrapperLocal(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[testModel] != "" {
-			testModel = modelMap[testModel]
-		}
-	}
-
 	cache, err := model.GetUserCache(1)
 	if err != nil {
 		return err, nil
@@ -97,7 +86,14 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 
 	middleware.SetupContextForSelectedChannel(c, channel, testModel)
 
-	meta := relaycommon.GenRelayInfo(c)
+	info := relaycommon.GenRelayInfo(c)
+
+	err = helper.ModelMappedHelper(c, info)
+	if err != nil {
+		return err, nil
+	}
+	testModel = info.UpstreamModelName
+
 	apiType, _ := constant.ChannelType2APIType(channel.Type)
 	adaptor := relay.GetAdaptor(apiType)
 	if adaptor == nil {
@@ -105,12 +101,11 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	}
 
 	request := buildTestRequest(testModel)
-	meta.UpstreamModelName = testModel
-	common.SysLog(fmt.Sprintf("testing channel %d with model %s , meta %v ", channel.Id, testModel, meta))
+	common.SysLog(fmt.Sprintf("testing channel %d with model %s , info %v ", channel.Id, testModel, info))
 
-	adaptor.Init(meta)
+	adaptor.Init(info)
 
-	convertedRequest, err := adaptor.ConvertRequest(c, meta, request)
+	convertedRequest, err := adaptor.ConvertRequest(c, info, request)
 	if err != nil {
 		return err, nil
 	}
@@ -120,7 +115,7 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	}
 	requestBody := bytes.NewBuffer(jsonData)
 	c.Request.Body = io.NopCloser(requestBody)
-	resp, err := adaptor.DoRequest(c, meta, requestBody)
+	resp, err := adaptor.DoRequest(c, info, requestBody)
 	if err != nil {
 		return err, nil
 	}
@@ -132,7 +127,7 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 			return fmt.Errorf("status code %d: %s", httpResp.StatusCode, err.Error.Message), err
 		}
 	}
-	usageA, respErr := adaptor.DoResponse(c, httpResp, meta)
+	usageA, respErr := adaptor.DoResponse(c, httpResp, info)
 	if respErr != nil {
 		return fmt.Errorf("%s", respErr.Error.Message), respErr
 	}
@@ -145,29 +140,27 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	if err != nil {
 		return err, nil
 	}
-	modelPrice, usePrice := common.GetModelPrice(testModel, false)
-	modelRatio, success := common.GetModelRatio(testModel)
-	if !success {
-		return fmt.Errorf("模型 %s 倍率未设置", testModel), nil
+	info.PromptTokens = usage.PromptTokens
+	priceData, err := helper.ModelPriceHelper(c, info, usage.PromptTokens, int(request.MaxTokens))
+	if err != nil {
+		return err, nil
 	}
-	completionRatio := common.GetCompletionRatio(testModel)
-	ratio := modelRatio
 	quota := 0
-	if !usePrice {
-		quota = usage.PromptTokens + int(math.Round(float64(usage.CompletionTokens)*completionRatio))
-		quota = int(math.Round(float64(quota) * ratio))
-		if ratio != 0 && quota <= 0 {
+	if !priceData.UsePrice {
+		quota = usage.PromptTokens + int(math.Round(float64(usage.CompletionTokens)*priceData.CompletionRatio))
+		quota = int(math.Round(float64(quota) * priceData.ModelRatio))
+		if priceData.ModelRatio != 0 && quota <= 0 {
 			quota = 1
 		}
 	} else {
-		quota = int(modelPrice * common.QuotaPerUnit)
+		quota = int(priceData.ModelPrice * common.QuotaPerUnit)
 	}
 	tok := time.Now()
 	milliseconds := tok.Sub(tik).Milliseconds()
 	consumedTime := float64(milliseconds) / 1000.0
-	other := service.GenerateTextOtherInfo(c, meta, modelRatio, 1, completionRatio, modelPrice)
-	model.RecordConsumeLog(c, 1, channel.Id, usage.PromptTokens, usage.CompletionTokens, testModel, "模型测试",
-		quota, "模型测试", 0, quota, int(consumedTime), false, "default", other)
+	other := service.GenerateTextOtherInfo(c, info, priceData.ModelRatio, priceData.GroupRatio, priceData.CompletionRatio, 0, 0.0, priceData.ModelPrice)
+	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)
 	common.SysLog(fmt.Sprintf("testing channel #%d, response: \n%s", channel.Id, string(respBody)))
 	return nil, nil
 }

controller/misc.go

@@ -7,6 +7,7 @@ import (
 	"one-api/common"
 	"one-api/model"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -66,7 +67,8 @@ func GetStatus(c *gin.Context) {
 			"enable_online_topup":      setting.PayAddress != "" && setting.EpayId != "" && setting.EpayKey != "",
 			"mj_notify_enabled":        setting.MjNotifyEnabled,
 			"chats":                    setting.Chats,
-			"demo_site_enabled":        setting.DemoSiteEnabled,
+			"demo_site_enabled":        operation_setting.DemoSiteEnabled,
+			"self_use_mode_enabled":    operation_setting.SelfUseModeEnabled,
 		},
 	})
 	return

controller/pricing.go

@@ -2,9 +2,9 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
-	"one-api/common"
 	"one-api/model"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 )
 
 func GetPricing(c *gin.Context) {
@@ -40,7 +40,7 @@ func GetPricing(c *gin.Context) {
 }
 
 func ResetModelRatio(c *gin.Context) {
-	defaultStr := common.DefaultModelRatio2JSONString()
+	defaultStr := operation_setting.DefaultModelRatio2JSONString()
 	err := model.UpdateOption("ModelRatio", defaultStr)
 	if err != nil {
 		c.JSON(200, gin.H{
@@ -49,7 +49,7 @@ func ResetModelRatio(c *gin.Context) {
 		})
 		return
 	}
-	err = common.UpdateModelRatioByJSONString(defaultStr)
+	err = operation_setting.UpdateModelRatioByJSONString(defaultStr)
 	if err != nil {
 		c.JSON(200, gin.H{
 			"success": false,

controller/relay.go

@@ -16,6 +16,7 @@ import (
 	"one-api/relay"
 	"one-api/relay/constant"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 )
@@ -41,15 +42,6 @@ func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode
 	return err
 }
 
-func wsHandler(c *gin.Context, ws *websocket.Conn, relayMode int) *dto.OpenAIErrorWithStatusCode {
-	var err *dto.OpenAIErrorWithStatusCode
-	switch relayMode {
-	default:
-		err = relay.TextHelper(c)
-	}
-	return err
-}
-
 func Relay(c *gin.Context) {
 	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
 	requestId := c.GetString(common.RequestIdKey)
@@ -110,7 +102,7 @@ func WssRelay(c *gin.Context) {
 
 	if err != nil {
 		openaiErr := service.OpenAIErrorWrapper(err, "get_channel_failed", http.StatusInternalServerError)
-		service.WssError(c, ws, openaiErr.Error)
+		helper.WssError(c, ws, openaiErr.Error)
 		return
 	}
 
@@ -152,7 +144,7 @@ func WssRelay(c *gin.Context) {
 			openaiErr.Error.Message = "当前分组上游负载已饱和，请稍后再试"
 		}
 		openaiErr.Error.Message = common.MessageWithRequestId(openaiErr.Error.Message, requestId)
-		service.WssError(c, ws, openaiErr.Error)
+		helper.WssError(c, ws, openaiErr.Error)
 	}
 }
 

dto/openai_request.go

@@ -99,6 +99,7 @@ type Message struct {
 	Name                *string         `json:"name,omitempty"`
 	Prefix              *bool           `json:"prefix,omitempty"`
 	ReasoningContent    string          `json:"reasoning_content,omitempty"`
+	Reasoning           string          `json:"reasoning,omitempty"`
 	ToolCalls           json.RawMessage `json:"tool_calls,omitempty"`
 	ToolCallId          string          `json:"tool_call_id,omitempty"`
 	parsedContent       []MediaContent

dto/openai_response.go

@@ -64,6 +64,7 @@ type ChatCompletionsStreamResponseChoice struct {
 type ChatCompletionsStreamResponseChoiceDelta struct {
 	Content          *string            `json:"content,omitempty"`
 	ReasoningContent *string            `json:"reasoning_content,omitempty"`
+	Reasoning        *string            `json:"reasoning,omitempty"`
 	Role             string             `json:"role,omitempty"`
 	ToolCalls        []ToolCallResponse `json:"tool_calls,omitempty"`
 }
@@ -80,14 +81,18 @@ func (c *ChatCompletionsStreamResponseChoiceDelta) GetContentString() string {
 }
 
 func (c *ChatCompletionsStreamResponseChoiceDelta) GetReasoningContent() string {
-	if c.ReasoningContent == nil {
+	if c.ReasoningContent == nil && c.Reasoning == nil {
 		return ""
 	}
-	return *c.ReasoningContent
+	if c.ReasoningContent != nil {
+		return *c.ReasoningContent
+	}
+	return *c.Reasoning
 }
 
 func (c *ChatCompletionsStreamResponseChoiceDelta) SetReasoningContent(s string) {
 	c.ReasoningContent = &s
+	c.Reasoning = &s
 }
 
 type ToolCallResponse struct {

middleware/model-rate-limit.go

@@ -51,7 +51,7 @@ func checkRedisRateLimit(ctx context.Context, rdb *redis.Client, key string, max
 	// 如果在时间窗口内已达到限制，拒绝请求
 	subTime := nowTime.Sub(oldTime).Seconds()
 	if int64(subTime) < duration {
-		rdb.Expire(ctx, key, common.RateLimitKeyExpirationDuration)
+		rdb.Expire(ctx, key, time.Duration(setting.ModelRequestRateLimitDurationMinutes)*time.Minute)
 		return false, nil
 	}
 
@@ -68,7 +68,7 @@ func recordRedisRequest(ctx context.Context, rdb *redis.Client, key string, maxC
 	now := time.Now().Format(timeFormat)
 	rdb.LPush(ctx, key, now)
 	rdb.LTrim(ctx, key, 0, int64(maxCount-1))
-	rdb.Expire(ctx, key, common.RateLimitKeyExpirationDuration)
+	rdb.Expire(ctx, key, time.Duration(setting.ModelRequestRateLimitDurationMinutes)*time.Minute)
 }
 
 // Redis限流处理器
@@ -118,7 +118,7 @@ func redisRateLimitHandler(duration int64, totalMaxCount, successMaxCount int) g
 
 // 内存限流处理器
 func memoryRateLimitHandler(duration int64, totalMaxCount, successMaxCount int) gin.HandlerFunc {
-	inMemoryRateLimiter.Init(common.RateLimitKeyExpirationDuration)
+	inMemoryRateLimiter.Init(time.Duration(setting.ModelRequestRateLimitDurationMinutes) * time.Minute)
 
 	return func(c *gin.Context) {
 		userId := strconv.Itoa(c.GetInt("id"))
@@ -153,20 +153,23 @@ func memoryRateLimitHandler(duration int64, totalMaxCount, successMaxCount int)
 
 // ModelRequestRateLimit 模型请求限流中间件
 func ModelRequestRateLimit() func(c *gin.Context) {
-	// 如果未启用限流，直接放行
-	if !setting.ModelRequestRateLimitEnabled {
-		return defNext
-	}
+	return func(c *gin.Context) {
+		// 在每个请求时检查是否启用限流
+		if !setting.ModelRequestRateLimitEnabled {
+			c.Next()
+			return
+		}
 
-	// 计算限流参数
-	duration := int64(setting.ModelRequestRateLimitDurationMinutes * 60)
-	totalMaxCount := setting.ModelRequestRateLimitCount
-	successMaxCount := setting.ModelRequestRateLimitSuccessCount
+		// 计算限流参数
+		duration := int64(setting.ModelRequestRateLimitDurationMinutes * 60)
+		totalMaxCount := setting.ModelRequestRateLimitCount
+		successMaxCount := setting.ModelRequestRateLimitSuccessCount
 
-	// 根据存储类型选择限流处理器
-	if common.RedisEnabled {
-		return redisRateLimitHandler(duration, totalMaxCount, successMaxCount)
-	} else {
-		return memoryRateLimitHandler(duration, totalMaxCount, successMaxCount)
+		// 根据存储类型选择并执行限流处理器
+		if common.RedisEnabled {
+			redisRateLimitHandler(duration, totalMaxCount, successMaxCount)(c)
+		} else {
+			memoryRateLimitHandler(duration, totalMaxCount, successMaxCount)(c)
+		}
 	}
 }

model/channel.go

@@ -290,35 +290,42 @@ func (channel *Channel) Delete() error {
 
 var channelStatusLock sync.Mutex
 
-func UpdateChannelStatusById(id int, status int, reason string) {
+func UpdateChannelStatusById(id int, status int, reason string) bool {
 	if common.MemoryCacheEnabled {
 		channelStatusLock.Lock()
+		defer channelStatusLock.Unlock()
+
 		channelCache, _ := CacheGetChannel(id)
 		// 如果缓存渠道存在，且状态已是目标状态，直接返回
 		if channelCache != nil && channelCache.Status == status {
-			channelStatusLock.Unlock()
-			return
+			return false
 		}
 		// 如果缓存渠道不存在(说明已经被禁用)，且要设置的状态不为启用，直接返回
 		if channelCache == nil && status != common.ChannelStatusEnabled {
-			channelStatusLock.Unlock()
-			return
+			return false
 		}
 		CacheUpdateChannelStatus(id, status)
-		channelStatusLock.Unlock()
 	}
 	err := UpdateAbilityStatus(id, status == common.ChannelStatusEnabled)
 	if err != nil {
 		common.SysError("failed to update ability status: " + err.Error())
+		return false
 	}
 	channel, err := GetChannelById(id, true)
 	if err != nil {
 		// find channel by id error, directly update status
-		err = DB.Model(&Channel{}).Where("id = ?", id).Update("status", status).Error
-		if err != nil {
-			common.SysError("failed to update channel status: " + err.Error())
+		result := DB.Model(&Channel{}).Where("id = ?", id).Update("status", status)
+		if result.Error != nil {
+			common.SysError("failed to update channel status: " + result.Error.Error())
+			return false
+		}
+		if result.RowsAffected == 0 {
+			return false
 		}
 	} else {
+		if channel.Status == status {
+			return false
+		}
 		// find channel by id success, update status and other info
 		info := channel.GetOtherInfo()
 		info["status_reason"] = reason
@@ -328,9 +335,10 @@ func UpdateChannelStatusById(id int, status int, reason string) {
 		err = channel.Save()
 		if err != nil {
 			common.SysError("failed to update channel status: " + err.Error())
+			return false
 		}
 	}
-
+	return true
 }
 
 func EnableChannelByTag(tag string) error {

model/log.go

@@ -2,12 +2,13 @@ package model
 
 import (
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"one-api/common"
 	"os"
 	"strings"
 	"time"
 
+	"github.com/gin-gonic/gin"
+
 	"github.com/bytedance/gopkg/util/gopool"
 	"gorm.io/gorm"
 )
@@ -18,7 +19,7 @@ type Log struct {
 	CreatedAt        int64  `json:"created_at" gorm:"bigint;index:idx_created_at_id,priority:2;index:idx_created_at_type"`
 	Type             int    `json:"type" gorm:"index:idx_created_at_type"`
 	Content          string `json:"content"`
-	Username         string `json:"username" gorm:"index:index_username_model_name,priority:2;default:''"`
+	Username         string `json:"username" gorm:"index;index:index_username_model_name,priority:2;default:''"`
 	TokenName        string `json:"token_name" gorm:"index;default:''"`
 	ModelName        string `json:"model_name" gorm:"index;index:index_username_model_name,priority:1;default:''"`
 	Quota            int    `json:"quota" gorm:"default:0"`

model/option.go

@@ -4,6 +4,7 @@ import (
 	"one-api/common"
 	"one-api/setting"
 	"one-api/setting/config"
+	"one-api/setting/operation_setting"
 	"strconv"
 	"strings"
 	"time"
@@ -87,15 +88,16 @@ func InitOptionMap() {
 	common.OptionMap["QuotaForInviter"] = strconv.Itoa(common.QuotaForInviter)
 	common.OptionMap["QuotaForInvitee"] = strconv.Itoa(common.QuotaForInvitee)
 	common.OptionMap["QuotaRemindThreshold"] = strconv.Itoa(common.QuotaRemindThreshold)
-	common.OptionMap["ShouldPreConsumedQuota"] = strconv.Itoa(common.PreConsumedQuota)
+	common.OptionMap["PreConsumedQuota"] = strconv.Itoa(common.PreConsumedQuota)
 	common.OptionMap["ModelRequestRateLimitCount"] = strconv.Itoa(setting.ModelRequestRateLimitCount)
 	common.OptionMap["ModelRequestRateLimitDurationMinutes"] = strconv.Itoa(setting.ModelRequestRateLimitDurationMinutes)
 	common.OptionMap["ModelRequestRateLimitSuccessCount"] = strconv.Itoa(setting.ModelRequestRateLimitSuccessCount)
-	common.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
-	common.OptionMap["ModelPrice"] = common.ModelPrice2JSONString()
+	common.OptionMap["ModelRatio"] = operation_setting.ModelRatio2JSONString()
+	common.OptionMap["ModelPrice"] = operation_setting.ModelPrice2JSONString()
+	common.OptionMap["CacheRatio"] = operation_setting.CacheRatio2JSONString()
 	common.OptionMap["GroupRatio"] = setting.GroupRatio2JSONString()
 	common.OptionMap["UserUsableGroups"] = setting.UserUsableGroups2JSONString()
-	common.OptionMap["CompletionRatio"] = common.CompletionRatio2JSONString()
+	common.OptionMap["CompletionRatio"] = operation_setting.CompletionRatio2JSONString()
 	common.OptionMap["TopUpLink"] = common.TopUpLink
 	common.OptionMap["ChatLink"] = common.ChatLink
 	common.OptionMap["ChatLink2"] = common.ChatLink2
@@ -110,13 +112,14 @@ func InitOptionMap() {
 	common.OptionMap["MjForwardUrlEnabled"] = strconv.FormatBool(setting.MjForwardUrlEnabled)
 	common.OptionMap["MjActionCheckSuccessEnabled"] = strconv.FormatBool(setting.MjActionCheckSuccessEnabled)
 	common.OptionMap["CheckSensitiveEnabled"] = strconv.FormatBool(setting.CheckSensitiveEnabled)
-	common.OptionMap["DemoSiteEnabled"] = strconv.FormatBool(setting.DemoSiteEnabled)
+	common.OptionMap["DemoSiteEnabled"] = strconv.FormatBool(operation_setting.DemoSiteEnabled)
+	common.OptionMap["SelfUseModeEnabled"] = strconv.FormatBool(operation_setting.SelfUseModeEnabled)
 	common.OptionMap["ModelRequestRateLimitEnabled"] = strconv.FormatBool(setting.ModelRequestRateLimitEnabled)
 	common.OptionMap["CheckSensitiveOnPromptEnabled"] = strconv.FormatBool(setting.CheckSensitiveOnPromptEnabled)
 	common.OptionMap["StopOnSensitiveEnabled"] = strconv.FormatBool(setting.StopOnSensitiveEnabled)
 	common.OptionMap["SensitiveWords"] = setting.SensitiveWordsToString()
 	common.OptionMap["StreamCacheQueueLength"] = strconv.Itoa(setting.StreamCacheQueueLength)
-	common.OptionMap["AutomaticDisableKeywords"] = setting.AutomaticDisableKeywordsToString()
+	common.OptionMap["AutomaticDisableKeywords"] = operation_setting.AutomaticDisableKeywordsToString()
 
 	// 自动添加所有注册的模型配置
 	modelConfigs := config.GlobalConfig.ExportAllConfigs()
@@ -242,7 +245,9 @@ func updateOptionMap(key string, value string) (err error) {
 		case "CheckSensitiveEnabled":
 			setting.CheckSensitiveEnabled = boolValue
 		case "DemoSiteEnabled":
-			setting.DemoSiteEnabled = boolValue
+			operation_setting.DemoSiteEnabled = boolValue
+		case "SelfUseModeEnabled":
+			operation_setting.SelfUseModeEnabled = boolValue
 		case "CheckSensitiveOnPromptEnabled":
 			setting.CheckSensitiveOnPromptEnabled = boolValue
 		case "ModelRequestRateLimitEnabled":
@@ -325,7 +330,7 @@ func updateOptionMap(key string, value string) (err error) {
 		common.QuotaForInvitee, _ = strconv.Atoi(value)
 	case "QuotaRemindThreshold":
 		common.QuotaRemindThreshold, _ = strconv.Atoi(value)
-	case "ShouldPreConsumedQuota":
+	case "PreConsumedQuota":
 		common.PreConsumedQuota, _ = strconv.Atoi(value)
 	case "ModelRequestRateLimitCount":
 		setting.ModelRequestRateLimitCount, _ = strconv.Atoi(value)
@@ -340,15 +345,17 @@ func updateOptionMap(key string, value string) (err error) {
 	case "DataExportDefaultTime":
 		common.DataExportDefaultTime = value
 	case "ModelRatio":
-		err = common.UpdateModelRatioByJSONString(value)
+		err = operation_setting.UpdateModelRatioByJSONString(value)
 	case "GroupRatio":
 		err = setting.UpdateGroupRatioByJSONString(value)
 	case "UserUsableGroups":
 		err = setting.UpdateUserUsableGroupsByJSONString(value)
 	case "CompletionRatio":
-		err = common.UpdateCompletionRatioByJSONString(value)
+		err = operation_setting.UpdateCompletionRatioByJSONString(value)
 	case "ModelPrice":
-		err = common.UpdateModelPriceByJSONString(value)
+		err = operation_setting.UpdateModelPriceByJSONString(value)
+	case "CacheRatio":
+		err = operation_setting.UpdateCacheRatioByJSONString(value)
 	case "TopUpLink":
 		common.TopUpLink = value
 	case "ChatLink":
@@ -362,7 +369,7 @@ func updateOptionMap(key string, value string) (err error) {
 	case "SensitiveWords":
 		setting.SensitiveWordsFromString(value)
 	case "AutomaticDisableKeywords":
-		setting.AutomaticDisableKeywordsFromString(value)
+		operation_setting.AutomaticDisableKeywordsFromString(value)
 	case "StreamCacheQueueLength":
 		setting.StreamCacheQueueLength, _ = strconv.Atoi(value)
 	}

model/pricing.go

@@ -2,6 +2,7 @@ package model
 
 import (
 	"one-api/common"
+	"one-api/setting/operation_setting"
 	"sync"
 	"time"
 )
@@ -64,14 +65,14 @@ func updatePricing() {
 			ModelName:   model,
 			EnableGroup: groups,
 		}
-		modelPrice, findPrice := common.GetModelPrice(model, false)
+		modelPrice, findPrice := operation_setting.GetModelPrice(model, false)
 		if findPrice {
 			pricing.ModelPrice = modelPrice
 			pricing.QuotaType = 1
 		} else {
-			modelRatio, _ := common.GetModelRatio(model)
+			modelRatio, _ := operation_setting.GetModelRatio(model)
 			pricing.ModelRatio = modelRatio
-			pricing.CompletionRatio = common.GetCompletionRatio(model)
+			pricing.CompletionRatio = operation_setting.GetCompletionRatio(model)
 			pricing.QuotaType = 0
 		}
 		pricingMap = append(pricingMap, pricing)

relay/channel/ali/text.go

@@ -8,6 +8,7 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 )
@@ -153,7 +154,7 @@ func aliStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWith
 		}
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	lastResponseText := ""
 	c.Stream(func(w io.Writer) bool {
 		select {

relay/channel/aws/dto.go

@@ -14,7 +14,7 @@ type AwsClaudeRequest struct {
 	TopP             float64                `json:"top_p,omitempty"`
 	TopK             int                    `json:"top_k,omitempty"`
 	StopSequences    []string               `json:"stop_sequences,omitempty"`
-	Tools            []claude.Tool          `json:"tools,omitempty"`
+	Tools            any                    `json:"tools,omitempty"`
 	ToolChoice       any                    `json:"tool_choice,omitempty"`
 	Thinking         *claude.Thinking       `json:"thinking,omitempty"`
 }

relay/channel/aws/relay-aws.go

@@ -12,6 +12,7 @@ import (
 	relaymodel "one-api/dto"
 	"one-api/relay/channel/claude"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"time"
@@ -203,13 +204,13 @@ func awsStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 		}
 	})
 	if info.ShouldIncludeUsage {
-		response := service.GenerateFinalUsageResponse(id, createdTime, info.UpstreamModelName, usage)
-		err := service.ObjectData(c, response)
+		response := helper.GenerateFinalUsageResponse(id, createdTime, info.UpstreamModelName, usage)
+		err := helper.ObjectData(c, response)
 		if err != nil {
 			common.SysError("send final response failed: " + err.Error())
 		}
 	}
-	service.Done(c)
+	helper.Done(c)
 	if resp != nil {
 		err = resp.Body.Close()
 		if err != nil {

relay/channel/baidu/relay-baidu.go

@@ -11,6 +11,7 @@ import (
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"sync"
@@ -138,7 +139,7 @@ func baiduStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWi
 		}
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:

relay/channel/claude/dto.go

@@ -58,7 +58,7 @@ type ClaudeRequest struct {
 	TopK              int             `json:"top_k,omitempty"`
 	//ClaudeMetadata    `json:"metadata,omitempty"`
 	Stream     bool      `json:"stream,omitempty"`
-	Tools      []Tool    `json:"tools,omitempty"`
+	Tools      any       `json:"tools,omitempty"`
 	ToolChoice any       `json:"tool_choice,omitempty"`
 	Thinking   *Thinking `json:"thinking,omitempty"`
 }

relay/channel/claude/relay-claude.go

@@ -1,7 +1,6 @@
 package claude
 
 import (
-	"bufio"
 	"encoding/json"
 	"fmt"
 	"io"
@@ -9,6 +8,7 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting/model_setting"
 	"strings"
@@ -443,28 +443,18 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 	usage = &dto.Usage{}
 	responseText := ""
 	createdTime := common.GetTimestamp()
-	scanner := bufio.NewScanner(resp.Body)
-	scanner.Split(bufio.ScanLines)
-	service.SetEventStreamHeaders(c)
 
-	for scanner.Scan() {
-		data := scanner.Text()
-		info.SetFirstResponseTime()
-		if len(data) < 6 || !strings.HasPrefix(data, "data:") {
-			continue
-		}
-		data = strings.TrimPrefix(data, "data:")
-		data = strings.TrimSpace(data)
+	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
 		var claudeResponse ClaudeResponse
 		err := json.Unmarshal([]byte(data), &claudeResponse)
 		if err != nil {
 			common.SysError("error unmarshalling stream response: " + err.Error())
-			continue
+			return true
 		}
 
 		response, claudeUsage := StreamResponseClaude2OpenAI(requestMode, &claudeResponse)
 		if response == nil {
-			continue
+			return true
 		}
 		if requestMode == RequestModeCompletion {
 			responseText += claudeResponse.Completion
@@ -481,9 +471,9 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 				usage.CompletionTokens = claudeUsage.OutputTokens
 				usage.TotalTokens = claudeUsage.InputTokens + claudeUsage.OutputTokens
 			} else if claudeResponse.Type == "content_block_start" {
-
+				return true
 			} else {
-				continue
+				return true
 			}
 		}
 		//response.Id = responseId
@@ -491,11 +481,12 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 		response.Created = createdTime
 		response.Model = info.UpstreamModelName
 
-		err = service.ObjectData(c, response)
+		err = helper.ObjectData(c, response)
 		if err != nil {
 			common.LogError(c, "send_stream_response_failed: "+err.Error())
 		}
-	}
+		return true
+	})
 
 	if requestMode == RequestModeCompletion {
 		usage, _ = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
@@ -508,14 +499,14 @@ func ClaudeStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 		}
 	}
 	if info.ShouldIncludeUsage {
-		response := service.GenerateFinalUsageResponse(responseId, createdTime, info.UpstreamModelName, *usage)
-		err := service.ObjectData(c, response)
+		response := helper.GenerateFinalUsageResponse(responseId, createdTime, info.UpstreamModelName, *usage)
+		err := helper.ObjectData(c, response)
 		if err != nil {
 			common.SysError("send final response failed: " + err.Error())
 		}
 	}
-	service.Done(c)
-	resp.Body.Close()
+	helper.Done(c)
+	//resp.Body.Close()
 	return nil, usage
 }
 

relay/channel/cloudflare/relay_cloudflare.go

@@ -9,6 +9,7 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"time"
@@ -28,8 +29,8 @@ func cfStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 
-	service.SetEventStreamHeaders(c)
-	id := service.GetResponseID(c)
+	helper.SetEventStreamHeaders(c)
+	id := helper.GetResponseID(c)
 	var responseText string
 	isFirst := true
 
@@ -57,7 +58,7 @@ func cfStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 		}
 		response.Id = id
 		response.Model = info.UpstreamModelName
-		err = service.ObjectData(c, response)
+		err = helper.ObjectData(c, response)
 		if isFirst {
 			isFirst = false
 			info.FirstResponseTime = time.Now()
@@ -72,13 +73,13 @@ func cfStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rela
 	}
 	usage, _ := service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	if info.ShouldIncludeUsage {
-		response := service.GenerateFinalUsageResponse(id, info.StartTime.Unix(), info.UpstreamModelName, *usage)
-		err := service.ObjectData(c, response)
+		response := helper.GenerateFinalUsageResponse(id, info.StartTime.Unix(), info.UpstreamModelName, *usage)
+		err := helper.ObjectData(c, response)
 		if err != nil {
 			common.LogError(c, "error_rendering_final_usage_response: "+err.Error())
 		}
 	}
-	service.Done(c)
+	helper.Done(c)
 
 	err := resp.Body.Close()
 	if err != nil {
@@ -109,7 +110,7 @@ func cfHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo)
 	}
 	usage, _ := service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	response.Usage = *usage
-	response.Id = service.GetResponseID(c)
+	response.Id = helper.GetResponseID(c)
 	jsonResponse, err := json.Marshal(response)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/cohere/relay-cohere.go

@@ -10,6 +10,7 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"time"
@@ -103,7 +104,7 @@ func cohereStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.
 		}
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	isFirst := true
 	c.Stream(func(w io.Writer) bool {
 		select {

relay/channel/dify/relay-dify.go

@@ -10,6 +10,7 @@ import (
 	"one-api/constant"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 )
@@ -66,7 +67,7 @@ func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 
 	for scanner.Scan() {
 		data := scanner.Text()
@@ -92,7 +93,7 @@ func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 				responseText += openaiResponse.Choices[0].Delta.GetContentString()
 			}
 		}
-		err = service.ObjectData(c, openaiResponse)
+		err = helper.ObjectData(c, openaiResponse)
 		if err != nil {
 			common.SysError(err.Error())
 		}
@@ -100,7 +101,7 @@ func difyStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Re
 	if err := scanner.Err(); err != nil {
 		common.SysError("error reading stream: " + err.Error())
 	}
-	service.Done(c)
+	helper.Done(c)
 	err := resp.Body.Close()
 	if err != nil {
 		//return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil

relay/channel/gemini/relay-gemini.go

@@ -1,7 +1,6 @@
 package gemini
 
 import (
-	"bufio"
 	"encoding/json"
 	"fmt"
 	"io"
@@ -10,6 +9,7 @@ import (
 	"one-api/constant"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting/model_setting"
 	"strings"
@@ -429,10 +429,10 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *dto.OpenAITextResp
 
 func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) (*dto.ChatCompletionsStreamResponse, bool) {
 	choices := make([]dto.ChatCompletionsStreamResponseChoice, 0, len(geminiResponse.Candidates))
-	is_stop := false
+	isStop := false
 	for _, candidate := range geminiResponse.Candidates {
 		if candidate.FinishReason != nil && *candidate.FinishReason == "STOP" {
-			is_stop = true
+			isStop = true
 			candidate.FinishReason = nil
 		}
 		choice := dto.ChatCompletionsStreamResponseChoice{
@@ -482,9 +482,8 @@ func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) (*dto.C
 
 	var response dto.ChatCompletionsStreamResponse
 	response.Object = "chat.completion.chunk"
-	response.Model = "gemini"
 	response.Choices = choices
-	return &response, is_stop
+	return &response, isStop
 }
 
 func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
@@ -492,27 +491,16 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 	id := fmt.Sprintf("chatcmpl-%s", common.GetUUID())
 	createAt := common.GetTimestamp()
 	var usage = &dto.Usage{}
-	scanner := bufio.NewScanner(resp.Body)
-	scanner.Split(bufio.ScanLines)
 
-	service.SetEventStreamHeaders(c)
-	for scanner.Scan() {
-		data := scanner.Text()
-		info.SetFirstResponseTime()
-		data = strings.TrimSpace(data)
-		if !strings.HasPrefix(data, "data: ") {
-			continue
-		}
-		data = strings.TrimPrefix(data, "data: ")
-		data = strings.TrimSuffix(data, "\"")
+	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
 		var geminiResponse GeminiChatResponse
 		err := json.Unmarshal([]byte(data), &geminiResponse)
 		if err != nil {
 			common.LogError(c, "error unmarshalling stream response: "+err.Error())
-			continue
+			return false
 		}
 
-		response, is_stop := streamResponseGeminiChat2OpenAI(&geminiResponse)
+		response, isStop := streamResponseGeminiChat2OpenAI(&geminiResponse)
 		response.Id = id
 		response.Created = createAt
 		response.Model = info.UpstreamModelName
@@ -521,15 +509,16 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 			usage.PromptTokens = geminiResponse.UsageMetadata.PromptTokenCount
 			usage.CompletionTokens = geminiResponse.UsageMetadata.CandidatesTokenCount
 		}
-		err = service.ObjectData(c, response)
+		err = helper.ObjectData(c, response)
 		if err != nil {
 			common.LogError(c, err.Error())
 		}
-		if is_stop {
-			response := service.GenerateStopResponse(id, createAt, info.UpstreamModelName, constant.FinishReasonStop)
-			service.ObjectData(c, response)
+		if isStop {
+			response := helper.GenerateStopResponse(id, createAt, info.UpstreamModelName, constant.FinishReasonStop)
+			helper.ObjectData(c, response)
 		}
-	}
+		return true
+	})
 
 	var response *dto.ChatCompletionsStreamResponse
 
@@ -538,14 +527,14 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 	usage.CompletionTokenDetails.TextTokens = usage.CompletionTokens
 
 	if info.ShouldIncludeUsage {
-		response = service.GenerateFinalUsageResponse(id, createAt, info.UpstreamModelName, *usage)
-		err := service.ObjectData(c, response)
+		response = helper.GenerateFinalUsageResponse(id, createAt, info.UpstreamModelName, *usage)
+		err := helper.ObjectData(c, response)
 		if err != nil {
 			common.SysError("send final response failed: " + err.Error())
 		}
 	}
-	service.Done(c)
-	resp.Body.Close()
+	helper.Done(c)
+	//resp.Body.Close()
 	return nil, usage
 }
 

relay/channel/openai/relay-openai.go

@@ -1,7 +1,6 @@
 package openai
 
 import (
-	"bufio"
 	"bytes"
 	"encoding/json"
 	"fmt"
@@ -14,11 +13,10 @@ import (
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"os"
 	"strings"
-	"sync"
-	"time"
 
 	"github.com/bytedance/gopkg/util/gopool"
 	"github.com/gin-gonic/gin"
@@ -32,7 +30,7 @@ func sendStreamData(c *gin.Context, info *relaycommon.RelayInfo, data string, fo
 	}
 
 	if !forceFormat && !thinkToContent {
-		return service.StringData(c, data)
+		return helper.StringData(c, data)
 	}
 
 	var lastStreamResponse dto.ChatCompletionsStreamResponse
@@ -41,44 +39,68 @@ func sendStreamData(c *gin.Context, info *relaycommon.RelayInfo, data string, fo
 	}
 
 	if !thinkToContent {
-		return service.ObjectData(c, lastStreamResponse)
+		return helper.ObjectData(c, lastStreamResponse)
+	}
+
+	hasThinkingContent := false
+	hasContent := false
+	var thinkingContent strings.Builder
+	for _, choice := range lastStreamResponse.Choices {
+		if len(choice.Delta.GetReasoningContent()) > 0 {
+			hasThinkingContent = true
+			thinkingContent.WriteString(choice.Delta.GetReasoningContent())
+		}
+		if len(choice.Delta.GetContentString()) > 0 {
+			hasContent = true
+		}
 	}
 
 	// Handle think to content conversion
-	if info.IsFirstResponse {
-		response := lastStreamResponse.Copy()
-		for i := range response.Choices {
-			response.Choices[i].Delta.SetContentString("<think>\n")
-			response.Choices[i].Delta.SetReasoningContent("")
+	if info.ThinkingContentInfo.IsFirstThinkingContent {
+		if hasThinkingContent {
+			response := lastStreamResponse.Copy()
+			for i := range response.Choices {
+				// send `think` tag with thinking content
+				response.Choices[i].Delta.SetContentString("<think>\n" + thinkingContent.String())
+				response.Choices[i].Delta.ReasoningContent = nil
+				response.Choices[i].Delta.Reasoning = nil
+			}
+			info.ThinkingContentInfo.IsFirstThinkingContent = false
+			return helper.ObjectData(c, response)
 		}
-		service.ObjectData(c, response)
 	}
 
 	if lastStreamResponse.Choices == nil || len(lastStreamResponse.Choices) == 0 {
-		return service.ObjectData(c, lastStreamResponse)
+		return helper.ObjectData(c, lastStreamResponse)
 	}
 
 	// Process each choice
 	for i, choice := range lastStreamResponse.Choices {
 		// Handle transition from thinking to content
-		if len(choice.Delta.GetContentString()) > 0 && !info.SendLastReasoningResponse {
+		if hasContent && !info.ThinkingContentInfo.SendLastThinkingContent {
 			response := lastStreamResponse.Copy()
 			for j := range response.Choices {
-				response.Choices[j].Delta.SetContentString("\n</think>")
-				response.Choices[j].Delta.SetReasoningContent("")
+				response.Choices[j].Delta.SetContentString("\n</think>\n")
+				response.Choices[j].Delta.ReasoningContent = nil
+				response.Choices[j].Delta.Reasoning = nil
 			}
-			info.SendLastReasoningResponse = true
-			service.ObjectData(c, response)
+			info.ThinkingContentInfo.SendLastThinkingContent = true
+			helper.ObjectData(c, response)
 		}
 
 		// Convert reasoning content to regular content
 		if len(choice.Delta.GetReasoningContent()) > 0 {
 			lastStreamResponse.Choices[i].Delta.SetContentString(choice.Delta.GetReasoningContent())
-			lastStreamResponse.Choices[i].Delta.SetReasoningContent("")
+			lastStreamResponse.Choices[i].Delta.ReasoningContent = nil
+			lastStreamResponse.Choices[i].Delta.Reasoning = nil
+		} else if !hasThinkingContent && !hasContent {
+			// flush thinking content
+			lastStreamResponse.Choices[i].Delta.ReasoningContent = nil
+			lastStreamResponse.Choices[i].Delta.Reasoning = nil
 		}
 	}
 
-	return service.ObjectData(c, lastStreamResponse)
+	return helper.ObjectData(c, lastStreamResponse)
 }
 
 func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
@@ -108,64 +130,22 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	}
 
 	toolCount := 0
-	scanner := bufio.NewScanner(resp.Body)
-	scanner.Split(bufio.ScanLines)
 
-	service.SetEventStreamHeaders(c)
-	streamingTimeout := time.Duration(constant.StreamingTimeout) * time.Second
-	if strings.HasPrefix(info.UpstreamModelName, "o1") || strings.HasPrefix(info.UpstreamModelName, "o3") {
-		// twice timeout for o1 model
-		streamingTimeout *= 2
-	}
-	ticker := time.NewTicker(streamingTimeout)
-	defer ticker.Stop()
-
-	stopChan := make(chan bool)
-	defer close(stopChan)
 	var (
 		lastStreamData string
-		mu             sync.Mutex
 	)
-	gopool.Go(func() {
-		for scanner.Scan() {
-			//info.SetFirstResponseTime()
-			ticker.Reset(time.Duration(constant.StreamingTimeout) * time.Second)
-			data := scanner.Text()
-			if common.DebugEnabled {
-				println(data)
-			}
-			if len(data) < 6 { // ignore blank line or wrong format
-				continue
-			}
-			if data[:5] != "data:" && data[:6] != "[DONE]" {
-				continue
-			}
-			mu.Lock()
-			data = data[5:]
-			data = strings.TrimSpace(data)
-			if !strings.HasPrefix(data, "[DONE]") {
-				if lastStreamData != "" {
-					err := sendStreamData(c, info, lastStreamData, forceFormat, thinkToContent)
-					if err != nil {
-						common.LogError(c, "streaming error: "+err.Error())
-					}
-					info.SetFirstResponseTime()
-				}
-				lastStreamData = data
-				streamItems = append(streamItems, data)
-			}
-			mu.Unlock()
-		}
-		common.SafeSendBool(stopChan, true)
-	})
 
-	select {
-	case <-ticker.C:
-		// 超时处理逻辑
-		common.LogError(c, "streaming timeout")
-	case <-stopChan:
-		// 正常结束
-	}
+	helper.StreamScannerHandler(c, resp, info, func(data string) bool {
+		if lastStreamData != "" {
+			err := sendStreamData(c, info, lastStreamData, forceFormat, thinkToContent)
+			if err != nil {
+				common.LogError(c, "streaming error: "+err.Error())
+			}
+		}
+		lastStreamData = data
+		streamItems = append(streamItems, data)
+		return true
+	})
 
 	shouldSendLastResp := true
 	var lastStreamResponse dto.ChatCompletionsStreamResponse
@@ -210,7 +190,10 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 					//}
 					for _, choice := range streamResponse.Choices {
 						responseTextBuilder.WriteString(choice.Delta.GetContentString())
+
+						// handle both reasoning_content and reasoning
 						responseTextBuilder.WriteString(choice.Delta.GetReasoningContent())
+
 						if choice.Delta.ToolCalls != nil {
 							if len(choice.Delta.ToolCalls) > toolCount {
 								toolCount = len(choice.Delta.ToolCalls)
@@ -231,7 +214,7 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 				//}
 				for _, choice := range streamResponse.Choices {
 					responseTextBuilder.WriteString(choice.Delta.GetContentString())
-					responseTextBuilder.WriteString(choice.Delta.GetReasoningContent())
+					responseTextBuilder.WriteString(choice.Delta.GetReasoningContent()) // This will handle both reasoning_content and reasoning
 					if choice.Delta.ToolCalls != nil {
 						if len(choice.Delta.ToolCalls) > toolCount {
 							toolCount = len(choice.Delta.ToolCalls)
@@ -274,14 +257,14 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	}
 
 	if info.ShouldIncludeUsage && !containStreamUsage {
-		response := service.GenerateFinalUsageResponse(responseId, createAt, model, *usage)
+		response := helper.GenerateFinalUsageResponse(responseId, createAt, model, *usage)
 		response.SetSystemFingerprint(systemFingerprint)
-		service.ObjectData(c, response)
+		helper.ObjectData(c, response)
 	}
 
-	service.Done(c)
+	helper.Done(c)
 
-	resp.Body.Close()
+	//resp.Body.Close()
 	return nil, usage
 }
 
@@ -323,7 +306,7 @@ func OpenaiHandler(c *gin.Context, resp *http.Response, promptTokens int, model
 	if simpleResponse.Usage.TotalTokens == 0 || (simpleResponse.Usage.PromptTokens == 0 && simpleResponse.Usage.CompletionTokens == 0) {
 		completionTokens := 0
 		for _, choice := range simpleResponse.Choices {
-			ctkm, _ := service.CountTextToken(choice.Message.StringContent()+choice.Message.ReasoningContent, model)
+			ctkm, _ := service.CountTextToken(choice.Message.StringContent()+choice.Message.ReasoningContent+choice.Message.Reasoning, model)
 			completionTokens += ctkm
 		}
 		simpleResponse.Usage = dto.Usage{
@@ -512,7 +495,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 				localUsage.InputTokenDetails.TextTokens += textToken
 				localUsage.InputTokenDetails.AudioTokens += audioToken
 
-				err = service.WssString(c, targetConn, string(message))
+				err = helper.WssString(c, targetConn, string(message))
 				if err != nil {
 					errChan <- fmt.Errorf("error writing to target: %v", err)
 					return
@@ -618,7 +601,7 @@ func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.Op
 					localUsage.OutputTokenDetails.AudioTokens += audioToken
 				}
 
-				err = service.WssString(c, clientConn, string(message))
+				err = helper.WssString(c, clientConn, string(message))
 				if err != nil {
 					errChan <- fmt.Errorf("error writing to client: %v", err)
 					return

relay/channel/palm/relay-palm.go

@@ -9,6 +9,7 @@ import (
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 )
 
@@ -112,7 +113,7 @@ func palmStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWit
 		dataChan <- string(jsonResponse)
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:

relay/channel/tencent/relay-tencent.go

@@ -14,6 +14,7 @@ import (
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strconv"
 	"strings"
@@ -91,7 +92,7 @@ func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 
 	for scanner.Scan() {
 		data := scanner.Text()
@@ -112,7 +113,7 @@ func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 			responseText += response.Choices[0].Delta.GetContentString()
 		}
 
-		err = service.ObjectData(c, response)
+		err = helper.ObjectData(c, response)
 		if err != nil {
 			common.SysError(err.Error())
 		}
@@ -122,7 +123,7 @@ func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 		common.SysError("error reading stream: " + err.Error())
 	}
 
-	service.Done(c)
+	helper.Done(c)
 
 	err := resp.Body.Close()
 	if err != nil {

relay/channel/vertex/adaptor.go

@@ -5,7 +5,6 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/jinzhu/copier"
 	"io"
 	"net/http"
 	"one-api/dto"
@@ -28,6 +27,7 @@ var claudeModelMap = map[string]string{
 	"claude-3-opus-20240229":     "claude-3-opus@20240229",
 	"claude-3-haiku-20240307":    "claude-3-haiku@20240307",
 	"claude-3-5-sonnet-20240620": "claude-3-5-sonnet@20240620",
+	"claude-3-5-sonnet-20241022": "claude-3-5-sonnet-v2@20241022",
 	"claude-3-7-sonnet-20250219": "claude-3-7-sonnet@20250219",
 }
 
@@ -86,15 +86,16 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		} else {
 			suffix = "rawPredict"
 		}
+		model := info.UpstreamModelName
 		if v, ok := claudeModelMap[info.UpstreamModelName]; ok {
-			info.UpstreamModelName = v
+			model = v
 		}
 		return fmt.Sprintf(
 			"https://%s-aiplatform.googleapis.com/v1/projects/%s/locations/%s/publishers/anthropic/models/%s:%s",
 			region,
 			adc.ProjectID,
 			region,
-			info.UpstreamModelName,
+			model,
 			suffix,
 		), nil
 	} else if a.RequestMode == RequestModeLlama {
@@ -127,13 +128,9 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, info *relaycommon.RelayInfo, re
 		if err != nil {
 			return nil, err
 		}
-		vertexClaudeReq := &VertexAIClaudeRequest{
-			AnthropicVersion: anthropicVersion,
-		}
-		if err = copier.Copy(vertexClaudeReq, claudeReq); err != nil {
-			return nil, errors.New("failed to copy claude request")
-		}
+		vertexClaudeReq := copyRequest(claudeReq, anthropicVersion)
 		c.Set("request_model", claudeReq.Model)
+		info.UpstreamModelName = claudeReq.Model
 		return vertexClaudeReq, nil
 	} else if a.RequestMode == RequestModeGemini {
 		geminiRequest, err := gemini.CovertGemini2OpenAI(*request)

relay/channel/vertex/dto.go

@@ -1,17 +1,37 @@
 package vertex
 
-import "one-api/relay/channel/claude"
+import (
+	"one-api/relay/channel/claude"
+)
 
 type VertexAIClaudeRequest struct {
 	AnthropicVersion string                 `json:"anthropic_version"`
 	Messages         []claude.ClaudeMessage `json:"messages"`
-	System           string                 `json:"system,omitempty"`
-	MaxTokens        int                    `json:"max_tokens,omitempty"`
+	System           any                    `json:"system,omitempty"`
+	MaxTokens        uint                   `json:"max_tokens,omitempty"`
 	StopSequences    []string               `json:"stop_sequences,omitempty"`
 	Stream           bool                   `json:"stream,omitempty"`
 	Temperature      *float64               `json:"temperature,omitempty"`
 	TopP             float64                `json:"top_p,omitempty"`
 	TopK             int                    `json:"top_k,omitempty"`
-	Tools            []claude.Tool          `json:"tools,omitempty"`
+	Tools            any                    `json:"tools,omitempty"`
 	ToolChoice       any                    `json:"tool_choice,omitempty"`
+	Thinking         *claude.Thinking       `json:"thinking,omitempty"`
+}
+
+func copyRequest(req *claude.ClaudeRequest, version string) *VertexAIClaudeRequest {
+	return &VertexAIClaudeRequest{
+		AnthropicVersion: version,
+		System:           req.System,
+		Messages:         req.Messages,
+		MaxTokens:        req.MaxTokens,
+		Stream:           req.Stream,
+		Temperature:      req.Temperature,
+		TopP:             req.TopP,
+		TopK:             req.TopK,
+		StopSequences:    req.StopSequences,
+		Tools:            req.Tools,
+		ToolChoice:       req.ToolChoice,
+		Thinking:         req.Thinking,
+	}
 }

relay/channel/xunfei/relay-xunfei.go

@@ -14,6 +14,7 @@ import (
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"time"
@@ -132,7 +133,7 @@ func xunfeiStreamHandler(c *gin.Context, textRequest dto.GeneralOpenAIRequest, a
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "make xunfei request err", http.StatusInternalServerError), nil
 	}
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	var usage dto.Usage
 	c.Stream(func(w io.Writer) bool {
 		select {

relay/channel/zhipu/relay-zhipu.go

@@ -10,6 +10,7 @@ import (
 	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"sync"
@@ -177,7 +178,7 @@ func zhipuStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWi
 		}
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:

relay/channel/zhipu_4v/relay-zhipu_v4.go

@@ -10,6 +10,7 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
+	"one-api/relay/helper"
 	"one-api/service"
 	"strings"
 	"sync"
@@ -197,7 +198,7 @@ func zhipuStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWi
 		}
 		stopChan <- true
 	}()
-	service.SetEventStreamHeaders(c)
+	helper.SetEventStreamHeaders(c)
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:

relay/common/relay_info.go

@@ -12,25 +12,30 @@ import (
 	"github.com/gorilla/websocket"
 )
 
+type ThinkingContentInfo struct {
+	IsFirstThinkingContent  bool
+	SendLastThinkingContent bool
+}
+
 type RelayInfo struct {
-	ChannelType               int
-	ChannelId                 int
-	TokenId                   int
-	TokenKey                  string
-	UserId                    int
-	Group                     string
-	TokenUnlimited            bool
-	StartTime                 time.Time
-	FirstResponseTime         time.Time
-	IsFirstResponse           bool
-	SendLastReasoningResponse bool
-	ApiType                   int
-	IsStream                  bool
-	IsPlayground              bool
-	UsePrice                  bool
-	RelayMode                 int
-	UpstreamModelName         string
-	OriginModelName           string
+	ChannelType       int
+	ChannelId         int
+	TokenId           int
+	TokenKey          string
+	UserId            int
+	Group             string
+	TokenUnlimited    bool
+	StartTime         time.Time
+	FirstResponseTime time.Time
+	isFirstResponse   bool
+	//SendLastReasoningResponse bool
+	ApiType           int
+	IsStream          bool
+	IsPlayground      bool
+	UsePrice          bool
+	RelayMode         int
+	UpstreamModelName string
+	OriginModelName   string
 	//RecodeModelName      string
 	RequestURLPath       string
 	ApiVersion           string
@@ -53,6 +58,7 @@ type RelayInfo struct {
 	UserSetting          map[string]interface{}
 	UserEmail            string
 	UserQuota            int
+	ThinkingContentInfo
 }
 
 // 定义支持流式选项的通道类型
@@ -95,7 +101,7 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 		UserQuota:         c.GetInt(constant.ContextKeyUserQuota),
 		UserSetting:       c.GetStringMap(constant.ContextKeyUserSetting),
 		UserEmail:         c.GetString(constant.ContextKeyUserEmail),
-		IsFirstResponse:   true,
+		isFirstResponse:   true,
 		RelayMode:         relayconstant.Path2RelayMode(c.Request.URL.Path),
 		BaseUrl:           c.GetString("base_url"),
 		RequestURLPath:    c.Request.URL.String(),
@@ -117,6 +123,10 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 		ApiKey:         strings.TrimPrefix(c.Request.Header.Get("Authorization"), "Bearer "),
 		Organization:   c.GetString("channel_organization"),
 		ChannelSetting: channelSetting,
+		ThinkingContentInfo: ThinkingContentInfo{
+			IsFirstThinkingContent:  true,
+			SendLastThinkingContent: false,
+		},
 	}
 	if strings.HasPrefix(c.Request.URL.Path, "/pg") {
 		info.IsPlayground = true
@@ -147,9 +157,9 @@ func (info *RelayInfo) SetIsStream(isStream bool) {
 }
 
 func (info *RelayInfo) SetFirstResponseTime() {
-	if info.IsFirstResponse {
+	if info.isFirstResponse {
 		info.FirstResponseTime = time.Now()
-		info.IsFirstResponse = false
+		info.isFirstResponse = false
 	}
 }
 

relay/helper/common.go

@@ -1,4 +1,4 @@
-package service
+package helper
 
 import (
 	"encoding/json"

relay/helper/price.go

@@ -6,31 +6,42 @@ import (
 	"one-api/common"
 	relaycommon "one-api/relay/common"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 )
 
 type PriceData struct {
 	ModelPrice             float64
 	ModelRatio             float64
+	CompletionRatio        float64
+	CacheRatio             float64
 	GroupRatio             float64
 	UsePrice               bool
 	ShouldPreConsumedQuota int
 }
 
 func ModelPriceHelper(c *gin.Context, info *relaycommon.RelayInfo, promptTokens int, maxTokens int) (PriceData, error) {
-	modelPrice, usePrice := common.GetModelPrice(info.OriginModelName, false)
+	modelPrice, usePrice := operation_setting.GetModelPrice(info.OriginModelName, false)
 	groupRatio := setting.GetGroupRatio(info.Group)
 	var preConsumedQuota int
 	var modelRatio float64
+	var completionRatio float64
+	var cacheRatio float64
 	if !usePrice {
 		preConsumedTokens := common.PreConsumedQuota
 		if maxTokens != 0 {
 			preConsumedTokens = promptTokens + maxTokens
 		}
 		var success bool
-		modelRatio, success = common.GetModelRatio(info.OriginModelName)
+		modelRatio, success = operation_setting.GetModelRatio(info.OriginModelName)
 		if !success {
-			return PriceData{}, fmt.Errorf("model %s ratio not found", info.OriginModelName)
+			if info.UserId == 1 {
+				return PriceData{}, fmt.Errorf("模型 %s 倍率或价格未配置，请设置或开始自用模式；Model %s ratio or price not set, please set or start self-use mode", info.OriginModelName, info.OriginModelName)
+			} else {
+				return PriceData{}, fmt.Errorf("模型 %s 倍率或价格未配置, 请联系管理员设置；Model %s ratio or price not set, please contact administrator to set", info.OriginModelName, info.OriginModelName)
+			}
 		}
+		completionRatio = operation_setting.GetCompletionRatio(info.OriginModelName)
+		cacheRatio, _ = operation_setting.GetCacheRatio(info.OriginModelName)
 		ratio := modelRatio * groupRatio
 		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
 	} else {
@@ -39,8 +50,10 @@ func ModelPriceHelper(c *gin.Context, info *relaycommon.RelayInfo, promptTokens
 	return PriceData{
 		ModelPrice:             modelPrice,
 		ModelRatio:             modelRatio,
+		CompletionRatio:        completionRatio,
 		GroupRatio:             groupRatio,
 		UsePrice:               usePrice,
+		CacheRatio:             cacheRatio,
 		ShouldPreConsumedQuota: preConsumedQuota,
 	}, nil
 }

relay/helper/stream_scanner.go

@@ -0,0 +1,91 @@
+package helper
+
+import (
+	"bufio"
+	"context"
+	"io"
+	"net/http"
+	"one-api/common"
+	"one-api/constant"
+	relaycommon "one-api/relay/common"
+	"strings"
+	"time"
+
+	"github.com/gin-gonic/gin"
+)
+
+func StreamScannerHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, dataHandler func(data string) bool) {
+
+	if resp == nil {
+		return
+	}
+
+	defer resp.Body.Close()
+
+	streamingTimeout := time.Duration(constant.StreamingTimeout) * time.Second
+	if strings.HasPrefix(info.UpstreamModelName, "o1") || strings.HasPrefix(info.UpstreamModelName, "o3") {
+		// twice timeout for thinking model
+		streamingTimeout *= 2
+	}
+
+	var (
+		stopChan = make(chan bool, 2)
+		scanner  = bufio.NewScanner(resp.Body)
+		ticker   = time.NewTicker(streamingTimeout)
+	)
+
+	defer func() {
+		ticker.Stop()
+		close(stopChan)
+	}()
+
+	scanner.Split(bufio.ScanLines)
+	SetEventStreamHeaders(c)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	ctx = context.WithValue(ctx, "stop_chan", stopChan)
+	common.RelayCtxGo(ctx, func() {
+		for scanner.Scan() {
+			ticker.Reset(streamingTimeout)
+			data := scanner.Text()
+			if common.DebugEnabled {
+				println(data)
+			}
+
+			if len(data) < 6 {
+				continue
+			}
+			if data[:5] != "data:" && data[:6] != "[DONE]" {
+				continue
+			}
+			data = data[5:]
+			data = strings.TrimLeft(data, " ")
+			data = strings.TrimSuffix(data, "\"")
+			if !strings.HasPrefix(data, "[DONE]") {
+				info.SetFirstResponseTime()
+				success := dataHandler(data)
+				if !success {
+					break
+				}
+			}
+		}
+
+		if err := scanner.Err(); err != nil {
+			if err != io.EOF {
+				common.LogError(c, "scanner error: "+err.Error())
+			}
+		}
+
+		common.SafeSendBool(stopChan, true)
+	})
+
+	select {
+	case <-ticker.C:
+		// 超时处理逻辑
+		common.LogError(c, "streaming timeout")
+	case <-stopChan:
+		// 正常结束
+	}
+}

relay/relay-audio.go

@@ -7,7 +7,6 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
-	"one-api/model"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
 	"one-api/relay/helper"
@@ -80,10 +79,6 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		return service.OpenAIErrorWrapperLocal(err, "model_price_error", http.StatusInternalServerError)
 	}
 
-	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
-	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "get_user_quota_failed", http.StatusInternalServerError)
-	}
 	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
 		return openaiErr

relay/relay-mj.go

@@ -15,6 +15,7 @@ import (
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 	"strconv"
 	"strings"
 	"time"
@@ -157,10 +158,10 @@ 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 := common.GetModelPrice(modelName, true)
+	modelPrice, success := operation_setting.GetModelPrice(modelName, true)
 	// 如果没有配置价格，则使用默认价格
 	if !success {
-		defaultPrice, ok := common.GetDefaultModelRatioMap()[modelName]
+		defaultPrice, ok := operation_setting.GetDefaultModelRatioMap()[modelName]
 		if !ok {
 			modelPrice = 0.1
 		} else {
@@ -463,10 +464,10 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 	fullRequestURL := fmt.Sprintf("%s%s", baseURL, requestURL)
 
 	modelName := service.CoverActionToModelName(midjRequest.Action)
-	modelPrice, success := common.GetModelPrice(modelName, true)
+	modelPrice, success := operation_setting.GetModelPrice(modelName, true)
 	// 如果没有配置价格，则使用默认价格
 	if !success {
-		defaultPrice, ok := common.GetDefaultModelRatioMap()[modelName]
+		defaultPrice, ok := operation_setting.GetDefaultModelRatioMap()[modelName]
 		if !ok {
 			modelPrice = 0.1
 		} else {

relay/relay-text.go

@@ -110,6 +110,7 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	if err != nil {
 		return service.OpenAIErrorWrapperLocal(err, "model_price_error", http.StatusInternalServerError)
 	}
+
 	// pre-consume quota 预消耗配额
 	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
@@ -303,24 +304,26 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 			CompletionTokens: 0,
 			TotalTokens:      relayInfo.PromptTokens,
 		}
-		extraContent += "  ，（可能是请求出错）"
+		extraContent += "（可能是请求出错）"
 	}
 	useTimeSeconds := time.Now().Unix() - relayInfo.StartTime.Unix()
 	promptTokens := usage.PromptTokens
+	cacheTokens := usage.PromptTokensDetails.CachedTokens
 	completionTokens := usage.CompletionTokens
 	modelName := relayInfo.OriginModelName
 
 	tokenName := ctx.GetString("token_name")
-	completionRatio := common.GetCompletionRatio(modelName)
+	completionRatio := priceData.CompletionRatio
+	cacheRatio := priceData.CacheRatio
 	ratio := priceData.ModelRatio * priceData.GroupRatio
 	modelRatio := priceData.ModelRatio
 	groupRatio := priceData.GroupRatio
 	modelPrice := priceData.ModelPrice
-	usePrice := priceData.UsePrice
 
 	quota := 0
 	if !priceData.UsePrice {
-		quota = promptTokens + int(math.Round(float64(completionTokens)*completionRatio))
+		quota = (promptTokens - cacheTokens) + int(math.Round(float64(cacheTokens)*cacheRatio))
+		quota += int(math.Round(float64(completionTokens) * completionRatio))
 		quota = int(math.Round(float64(quota) * ratio))
 		if ratio != 0 && quota <= 0 {
 			quota = 1
@@ -329,8 +332,9 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 		quota = int(modelPrice * common.QuotaPerUnit * groupRatio)
 	}
 	totalTokens := promptTokens + completionTokens
+	
 	var logContent string
-	if !usePrice {
+	if !priceData.UsePrice {
 		logContent = fmt.Sprintf("模型倍率 %.2f，补全倍率 %.2f，分组倍率 %.2f", modelRatio, completionRatio, groupRatio)
 	} else {
 		logContent = fmt.Sprintf("模型价格 %.2f，分组倍率 %.2f", modelPrice, groupRatio)
@@ -371,7 +375,7 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 	if extraContent != "" {
 		logContent += ", " + extraContent
 	}
-	other := service.GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, modelPrice)
+	other := service.GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, cacheTokens, cacheRatio, modelPrice)
 	model.RecordConsumeLog(ctx, relayInfo.UserId, relayInfo.ChannelId, promptTokens, completionTokens, logModel,
 		tokenName, quota, logContent, relayInfo.TokenId, userQuota, int(useTimeSeconds), relayInfo.IsStream, relayInfo.Group, other)
 

relay/relay_task.go

@@ -16,6 +16,7 @@ import (
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 )
 
 /*
@@ -37,9 +38,9 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 	}
 
 	modelName := service.CoverTaskActionToModelName(platform, relayInfo.Action)
-	modelPrice, success := common.GetModelPrice(modelName, true)
+	modelPrice, success := operation_setting.GetModelPrice(modelName, true)
 	if !success {
-		defaultPrice, ok := common.GetDefaultModelRatioMap()[modelName]
+		defaultPrice, ok := operation_setting.GetDefaultModelRatioMap()[modelName]
 		if !ok {
 			modelPrice = 0.1
 		} else {

relay/websocket.go

@@ -11,6 +11,7 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/service"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 )
 
 func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWithStatusCode) {
@@ -39,7 +40,7 @@ func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWi
 		}
 	}
 	//relayInfo.UpstreamModelName = textRequest.Model
-	modelPrice, getModelPriceSuccess := common.GetModelPrice(relayInfo.UpstreamModelName, false)
+	modelPrice, getModelPriceSuccess := operation_setting.GetModelPrice(relayInfo.UpstreamModelName, false)
 	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 
 	var preConsumedQuota int
@@ -65,7 +66,7 @@ func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWi
 		//if realtimeEvent.Session.MaxResponseOutputTokens != 0 {
 		//	preConsumedTokens = promptTokens + int(realtimeEvent.Session.MaxResponseOutputTokens)
 		//}
-		modelRatio, _ = common.GetModelRatio(relayInfo.UpstreamModelName)
+		modelRatio, _ = operation_setting.GetModelRatio(relayInfo.UpstreamModelName)
 		ratio = modelRatio * groupRatio
 		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
 	} else {

service/channel.go

@@ -6,23 +6,31 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	"one-api/model"
-	"one-api/setting"
+	"one-api/setting/operation_setting"
 	"strings"
 )
 
+func formatNotifyType(channelId int, status int) string {
+	return fmt.Sprintf("%s_%d_%d", dto.NotifyTypeChannelUpdate, channelId, status)
+}
+
 // disable & notify
 func DisableChannel(channelId int, channelName string, reason string) {
-	model.UpdateChannelStatusById(channelId, common.ChannelStatusAutoDisabled, reason)
-	subject := fmt.Sprintf("通道「%s」（#%d）已被禁用", channelName, channelId)
-	content := fmt.Sprintf("通道「%s」（#%d）已被禁用，原因：%s", channelName, channelId, reason)
-	NotifyRootUser(subject, content, dto.NotifyTypeChannelUpdate)
+	success := model.UpdateChannelStatusById(channelId, common.ChannelStatusAutoDisabled, reason)
+	if success {
+		subject := fmt.Sprintf("通道「%s」（#%d）已被禁用", channelName, channelId)
+		content := fmt.Sprintf("通道「%s」（#%d）已被禁用，原因：%s", channelName, channelId, reason)
+		NotifyRootUser(formatNotifyType(channelId, common.ChannelStatusAutoDisabled), subject, content)
+	}
 }
 
 func EnableChannel(channelId int, channelName string) {
-	model.UpdateChannelStatusById(channelId, common.ChannelStatusEnabled, "")
-	subject := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
-	content := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
-	NotifyRootUser(subject, content, dto.NotifyTypeChannelUpdate)
+	success := model.UpdateChannelStatusById(channelId, common.ChannelStatusEnabled, "")
+	if success {
+		subject := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
+		content := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
+		NotifyRootUser(formatNotifyType(channelId, common.ChannelStatusEnabled), subject, content)
+	}
 }
 
 func ShouldDisableChannel(channelType int, err *dto.OpenAIErrorWithStatusCode) bool {
@@ -67,7 +75,7 @@ func ShouldDisableChannel(channelType int, err *dto.OpenAIErrorWithStatusCode) b
 	}
 
 	lowerMessage := strings.ToLower(err.Error.Message)
-	search, _ := AcSearch(lowerMessage, setting.AutomaticDisableKeywords, true)
+	search, _ := AcSearch(lowerMessage, operation_setting.AutomaticDisableKeywords, true)
 	if search {
 		return true
 	}

service/image.go

@@ -7,7 +7,9 @@ import (
 	"fmt"
 	"image"
 	"io"
+	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"strings"
 
 	"golang.org/x/image/webp"
@@ -23,7 +25,7 @@ func DecodeBase64ImageData(base64String string) (image.Config, string, string, e
 	decodedData, err := base64.StdEncoding.DecodeString(base64String)
 	if err != nil {
 		fmt.Println("Error: Failed to decode base64 string")
-		return image.Config{}, "", "", err
+		return image.Config{}, "", "", fmt.Errorf("failed to decode base64 string: %s", err.Error())
 	}
 
 	// 创建一个bytes.Buffer用于存储解码后的数据
@@ -61,20 +63,51 @@ func DecodeBase64FileData(base64String string) (string, string, error) {
 func GetImageFromUrl(url string) (mimeType string, data string, err error) {
 	resp, err := DoDownloadRequest(url)
 	if err != nil {
-		return "", "", err
-	}
-	if !strings.HasPrefix(resp.Header.Get("Content-Type"), "image/") {
-		return "", "", fmt.Errorf("invalid content type: %s, required image/*", resp.Header.Get("Content-Type"))
+		return "", "", fmt.Errorf("failed to download image: %w", err)
 	}
 	defer resp.Body.Close()
-	buffer := bytes.NewBuffer(nil)
-	_, err = buffer.ReadFrom(resp.Body)
-	if err != nil {
-		return
+
+	// Check HTTP status code
+	if resp.StatusCode != http.StatusOK {
+		return "", "", fmt.Errorf("failed to download image: HTTP %d", resp.StatusCode)
 	}
-	mimeType = resp.Header.Get("Content-Type")
+
+	contentType := resp.Header.Get("Content-Type")
+	if contentType != "application/octet-stream" && !strings.HasPrefix(contentType, "image/") {
+		return "", "", fmt.Errorf("invalid content type: %s, required image/*", contentType)
+	}
+	maxImageSize := int64(constant.MaxFileDownloadMB * 1024 * 1024)
+
+	// Check Content-Length if available
+	if resp.ContentLength > maxImageSize {
+		return "", "", fmt.Errorf("image size %d exceeds maximum allowed size of %d bytes", resp.ContentLength, maxImageSize)
+	}
+
+	// Use LimitReader to prevent reading oversized images
+	limitReader := io.LimitReader(resp.Body, maxImageSize)
+	buffer := &bytes.Buffer{}
+
+	written, err := io.Copy(buffer, limitReader)
+	if err != nil {
+		return "", "", fmt.Errorf("failed to read image data: %w", err)
+	}
+	if written >= maxImageSize {
+		return "", "", fmt.Errorf("image size exceeds maximum allowed size of %d bytes", maxImageSize)
+	}
+
 	data = base64.StdEncoding.EncodeToString(buffer.Bytes())
-	return
+	mimeType = contentType
+
+	// Handle application/octet-stream type
+	if mimeType == "application/octet-stream" {
+		_, format, _, err := DecodeBase64ImageData(data)
+		if err != nil {
+			return "", "", err
+		}
+		mimeType = "image/" + format
+	}
+
+	return mimeType, data, nil
 }
 
 func DecodeUrlImageData(imageUrl string) (image.Config, string, error) {
@@ -92,7 +125,7 @@ func DecodeUrlImageData(imageUrl string) (image.Config, string, error) {
 
 	mimeType := response.Header.Get("Content-Type")
 
-	if !strings.HasPrefix(mimeType, "image/") {
+	if mimeType != "application/octet-stream" && !strings.HasPrefix(mimeType, "image/") {
 		return image.Config{}, "", fmt.Errorf("invalid content type: %s, required image/*", mimeType)
 	}
 

service/log_info_generate.go

@@ -1,16 +1,20 @@
 package service
 
 import (
-	"github.com/gin-gonic/gin"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+
+	"github.com/gin-gonic/gin"
 )
 
-func GenerateTextOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelRatio, groupRatio, completionRatio, modelPrice float64) map[string]interface{} {
+func GenerateTextOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelRatio, groupRatio, completionRatio float64,
+	cacheTokens int, cacheRatio float64, modelPrice float64) map[string]interface{} {
 	other := make(map[string]interface{})
 	other["model_ratio"] = modelRatio
 	other["group_ratio"] = groupRatio
 	other["completion_ratio"] = completionRatio
+	other["cache_tokens"] = cacheTokens
+	other["cache_ratio"] = cacheRatio
 	other["model_price"] = modelPrice
 	other["frt"] = float64(relayInfo.FirstResponseTime.UnixMilli() - relayInfo.StartTime.UnixMilli())
 	if relayInfo.ReasoningEffort != "" {
@@ -27,7 +31,7 @@ func GenerateTextOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, m
 }
 
 func GenerateWssOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usage *dto.RealtimeUsage, modelRatio, groupRatio, completionRatio, audioRatio, audioCompletionRatio, modelPrice float64) map[string]interface{} {
-	info := GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, modelPrice)
+	info := GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, 0, 0.0, modelPrice)
 	info["ws"] = true
 	info["audio_input"] = usage.InputTokenDetails.AudioTokens
 	info["audio_output"] = usage.OutputTokenDetails.AudioTokens
@@ -39,7 +43,7 @@ func GenerateWssOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, us
 }
 
 func GenerateAudioOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usage *dto.Usage, modelRatio, groupRatio, completionRatio, audioRatio, audioCompletionRatio, modelPrice float64) map[string]interface{} {
-	info := GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, modelPrice)
+	info := GenerateTextOtherInfo(ctx, relayInfo, modelRatio, groupRatio, completionRatio, 0, 0.0, modelPrice)
 	info["audio"] = true
 	info["audio_input"] = usage.PromptTokensDetails.AudioTokens
 	info["audio_output"] = usage.CompletionTokenDetails.AudioTokens

service/quota.go

@@ -12,6 +12,7 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/relay/helper"
 	"one-api/setting"
+	"one-api/setting/operation_setting"
 	"strings"
 	"time"
 
@@ -38,9 +39,9 @@ func calculateAudioQuota(info QuotaInfo) int {
 		return int(info.ModelPrice * common.QuotaPerUnit * info.GroupRatio)
 	}
 
-	completionRatio := common.GetCompletionRatio(info.ModelName)
-	audioRatio := common.GetAudioRatio(info.ModelName)
-	audioCompletionRatio := common.GetAudioCompletionRatio(info.ModelName)
+	completionRatio := operation_setting.GetCompletionRatio(info.ModelName)
+	audioRatio := operation_setting.GetAudioRatio(info.ModelName)
+	audioCompletionRatio := operation_setting.GetAudioCompletionRatio(info.ModelName)
 	ratio := info.GroupRatio * info.ModelRatio
 
 	quota := info.InputDetails.TextTokens + int(math.Round(float64(info.OutputDetails.TextTokens)*completionRatio))
@@ -75,7 +76,7 @@ func PreWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usag
 	audioInputTokens := usage.InputTokenDetails.AudioTokens
 	audioOutTokens := usage.OutputTokenDetails.AudioTokens
 	groupRatio := setting.GetGroupRatio(relayInfo.Group)
-	modelRatio, _ := common.GetModelRatio(modelName)
+	modelRatio, _ := operation_setting.GetModelRatio(modelName)
 
 	quotaInfo := QuotaInfo{
 		InputDetails: TokenDetails{
@@ -122,9 +123,9 @@ func PostWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, mod
 	audioOutTokens := usage.OutputTokenDetails.AudioTokens
 
 	tokenName := ctx.GetString("token_name")
-	completionRatio := common.GetCompletionRatio(modelName)
-	audioRatio := common.GetAudioRatio(relayInfo.OriginModelName)
-	audioCompletionRatio := common.GetAudioCompletionRatio(modelName)
+	completionRatio := operation_setting.GetCompletionRatio(modelName)
+	audioRatio := operation_setting.GetAudioRatio(relayInfo.OriginModelName)
+	audioCompletionRatio := operation_setting.GetAudioCompletionRatio(modelName)
 
 	quotaInfo := QuotaInfo{
 		InputDetails: TokenDetails{
@@ -184,9 +185,9 @@ func PostAudioConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 	audioOutTokens := usage.CompletionTokenDetails.AudioTokens
 
 	tokenName := ctx.GetString("token_name")
-	completionRatio := common.GetCompletionRatio(relayInfo.OriginModelName)
-	audioRatio := common.GetAudioRatio(relayInfo.OriginModelName)
-	audioCompletionRatio := common.GetAudioCompletionRatio(relayInfo.OriginModelName)
+	completionRatio := operation_setting.GetCompletionRatio(relayInfo.OriginModelName)
+	audioRatio := operation_setting.GetAudioRatio(relayInfo.OriginModelName)
+	audioCompletionRatio := operation_setting.GetAudioCompletionRatio(relayInfo.OriginModelName)
 
 	modelRatio := priceData.ModelRatio
 	groupRatio := priceData.GroupRatio

service/token_counter.go

@@ -10,6 +10,7 @@ import (
 	"one-api/constant"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/setting/operation_setting"
 	"strings"
 	"unicode/utf8"
 
@@ -32,7 +33,7 @@ func InitTokenEncoders() {
 	if err != nil {
 		common.FatalLog(fmt.Sprintf("failed to get gpt-4o token encoder: %s", err.Error()))
 	}
-	for model, _ := range common.GetDefaultModelRatioMap() {
+	for model, _ := range operation_setting.GetDefaultModelRatioMap() {
 		if strings.HasPrefix(model, "gpt-3.5") {
 			tokenEncoderMap[model] = cl100TokenEncoder
 		} else if strings.HasPrefix(model, "gpt-4") {

service/user_notify.go

@@ -11,7 +11,10 @@ import (
 
 func NotifyRootUser(t string, subject string, content string) {
 	user := model.GetRootUser().ToBaseUser()
-	_ = NotifyUser(user.Id, user.Email, user.GetSetting(), dto.NewNotify(t, subject, content, nil))
+	err := NotifyUser(user.Id, user.Email, user.GetSetting(), dto.NewNotify(t, subject, content, nil))
+	if err != nil {
+		common.SysError(fmt.Sprintf("failed to notify root user: %s", err.Error()))
+	}
 }
 
 func NotifyUser(userId int, userEmail string, userSetting map[string]interface{}, data dto.Notify) error {

setting/operation_setting/cache_ratio.go

@@ -0,0 +1,77 @@
+package operation_setting
+
+import (
+	"encoding/json"
+	"one-api/common"
+	"sync"
+)
+
+var defaultCacheRatio = map[string]float64{
+	"gpt-4":                        0.5,
+	"o1-2024-12-17":                0.5,
+	"o1-preview-2024-09-12":        0.5,
+	"o1-mini-2024-09-12":           0.5,
+	"gpt-4o-2024-11-20":            0.5,
+	"gpt-4o-2024-08-06":            0.5,
+	"gpt-4o-mini-2024-07-18":       0.5,
+	"gpt-4o-realtime-preview":      0.5,
+	"gpt-4o-mini-realtime-preview": 0.5,
+	"deepseek-chat":                0.5,
+	"deepseek-reasoner":            0.5,
+	"deepseek-coder":               0.5,
+}
+
+var cacheRatioMap map[string]float64
+var cacheRatioMapMutex sync.RWMutex
+
+// GetCacheRatioMap returns the cache ratio map
+func GetCacheRatioMap() map[string]float64 {
+	cacheRatioMapMutex.Lock()
+	defer cacheRatioMapMutex.Unlock()
+	if cacheRatioMap == nil {
+		cacheRatioMap = defaultCacheRatio
+	}
+	return cacheRatioMap
+}
+
+// CacheRatio2JSONString converts the cache ratio map to a JSON string
+func CacheRatio2JSONString() string {
+	GetCacheRatioMap()
+	jsonBytes, err := json.Marshal(cacheRatioMap)
+	if err != nil {
+		common.SysError("error marshalling cache ratio: " + err.Error())
+	}
+	return string(jsonBytes)
+}
+
+// UpdateCacheRatioByJSONString updates the cache ratio map from a JSON string
+func UpdateCacheRatioByJSONString(jsonStr string) error {
+	cacheRatioMapMutex.Lock()
+	defer cacheRatioMapMutex.Unlock()
+	cacheRatioMap = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &cacheRatioMap)
+}
+
+// GetCacheRatio returns the cache ratio for a model
+func GetCacheRatio(name string) (float64, bool) {
+	GetCacheRatioMap()
+	ratio, ok := cacheRatioMap[name]
+	if !ok {
+		return 0.5, false // Default to 0.5 if not found
+	}
+	return ratio, true
+}
+
+// DefaultCacheRatio2JSONString converts the default cache ratio map to a JSON string
+func DefaultCacheRatio2JSONString() string {
+	jsonBytes, err := json.Marshal(defaultCacheRatio)
+	if err != nil {
+		common.SysError("error marshalling default cache ratio: " + err.Error())
+	}
+	return string(jsonBytes)
+}
+
+// GetDefaultCacheRatioMap returns the default cache ratio map
+func GetDefaultCacheRatioMap() map[string]float64 {
+	return defaultCacheRatio
+}

setting/operation_setting/model-ratio.go

@@ -1,7 +1,8 @@
-package common
+package operation_setting
 
 import (
 	"encoding/json"
+	"one-api/common"
 	"strings"
 	"sync"
 )
@@ -261,7 +262,7 @@ func ModelPrice2JSONString() string {
 	GetModelPriceMap()
 	jsonBytes, err := json.Marshal(modelPriceMap)
 	if err != nil {
-		SysError("error marshalling model price: " + err.Error())
+		common.SysError("error marshalling model price: " + err.Error())
 	}
 	return string(jsonBytes)
 }
@@ -285,7 +286,7 @@ func GetModelPrice(name string, printErr bool) (float64, bool) {
 	price, ok := modelPriceMap[name]
 	if !ok {
 		if printErr {
-			SysError("model price not found: " + name)
+			common.SysError("model price not found: " + name)
 		}
 		return -1, false
 	}
@@ -305,7 +306,7 @@ func ModelRatio2JSONString() string {
 	GetModelRatioMap()
 	jsonBytes, err := json.Marshal(modelRatioMap)
 	if err != nil {
-		SysError("error marshalling model ratio: " + err.Error())
+		common.SysError("error marshalling model ratio: " + err.Error())
 	}
 	return string(jsonBytes)
 }
@@ -324,8 +325,7 @@ func GetModelRatio(name string) (float64, bool) {
 	}
 	ratio, ok := modelRatioMap[name]
 	if !ok {
-		SysError("model ratio not found: " + name)
-		return 37.5, false
+		return 37.5, SelfUseModeEnabled
 	}
 	return ratio, true
 }
@@ -333,7 +333,7 @@ func GetModelRatio(name string) (float64, bool) {
 func DefaultModelRatio2JSONString() string {
 	jsonBytes, err := json.Marshal(defaultModelRatio)
 	if err != nil {
-		SysError("error marshalling model ratio: " + err.Error())
+		common.SysError("error marshalling model ratio: " + err.Error())
 	}
 	return string(jsonBytes)
 }
@@ -355,7 +355,7 @@ func CompletionRatio2JSONString() string {
 	GetCompletionRatioMap()
 	jsonBytes, err := json.Marshal(CompletionRatio)
 	if err != nil {
-		SysError("error marshalling completion ratio: " + err.Error())
+		common.SysError("error marshalling completion ratio: " + err.Error())
 	}
 	return string(jsonBytes)
 }

setting/operation_setting/operation_setting.go

@@ -1,8 +1,9 @@
-package setting
+package operation_setting
 
 import "strings"
 
 var DemoSiteEnabled = false
+var SelfUseModeEnabled = false
 
 var AutomaticDisableKeywords = []string{
 	"Your credit balance is too low",

web/src/App.js

@@ -30,6 +30,7 @@ import { useTranslation } from 'react-i18next';
 import { StatusContext } from './context/Status';
 import { setStatusData } from './helpers/data.js';
 import { API, showError } from './helpers';
+import PersonalSetting from './components/PersonalSetting.js';
 
 const Home = lazy(() => import('./pages/Home'));
 const Detail = lazy(() => import('./pages/Detail'));
@@ -177,6 +178,16 @@ function App() {
             </PrivateRoute>
           }
         />
+        <Route
+          path='/personal'
+          element={
+            <PrivateRoute>
+              <Suspense fallback={<Loading></Loading>}>
+                <PersonalSetting />
+              </Suspense>
+            </PrivateRoute>
+          }
+        />
         <Route
           path='/topup'
           element={

web/src/components/ChannelsTable.js

@@ -15,7 +15,7 @@ import {
   getQuotaPerUnit,
   renderGroup,
   renderNumberWithPoint,
-  renderQuota, renderQuotaWithPrompt
+  renderQuota, renderQuotaWithPrompt, stringToColor
 } from '../helpers/render';
 import {
   Button, Divider,
@@ -378,17 +378,15 @@ const ChannelsTable = () => {
                 >
                   {t('测试')}
                 </Button>
-                <Dropdown
-                  trigger="click"
-                  position="bottomRight"
-                  menu={modelMenuItems}  // 使用即时生成的菜单项
-                >
-                  <Button
-                    style={{ padding: '8px 4px' }}
-                    type="primary"
-                    icon={<IconTreeTriangleDown />}
-                  ></Button>
-                </Dropdown>
+                <Button
+                  style={{ padding: '8px 4px' }}
+                  type="primary"
+                  icon={<IconTreeTriangleDown />}
+                  onClick={() => {
+                    setCurrentTestChannel(record);
+                    setShowModelTestModal(true);
+                  }}
+                ></Button>
               </SplitButtonGroup>
               <Popconfirm
                 title={t('确定是否要删除此渠道？')}
@@ -522,6 +520,9 @@ const ChannelsTable = () => {
   const [enableTagMode, setEnableTagMode] = useState(false);
   const [showBatchSetTag, setShowBatchSetTag] = useState(false);
   const [batchSetTagValue, setBatchSetTagValue] = useState('');
+  const [showModelTestModal, setShowModelTestModal] = useState(false);
+  const [currentTestChannel, setCurrentTestChannel] = useState(null);
+  const [modelSearchKeyword, setModelSearchKeyword] = useState('');
 
 
   const removeRecord = (record) => {
@@ -1289,6 +1290,77 @@ const ChannelsTable = () => {
           onChange={(v) => setBatchSetTagValue(v)}
         />
       </Modal>
+      
+      {/* 模型测试弹窗 */}
+      <Modal
+        title={t('选择模型进行测试')}
+        visible={showModelTestModal && currentTestChannel !== null}
+        onCancel={() => {
+          setShowModelTestModal(false);
+          setModelSearchKeyword('');
+        }}
+        footer={null}
+        maskClosable={true}
+        centered={true}
+        width={600}
+      >
+        <div style={{ maxHeight: '500px', overflowY: 'auto', padding: '10px' }}>
+          {currentTestChannel && (
+            <div>
+              <Typography.Title heading={6} style={{ marginBottom: '16px' }}>
+                {t('渠道')}: {currentTestChannel.name}
+              </Typography.Title>
+              
+              {/* 搜索框 */}
+              <Input
+                placeholder={t('搜索模型...')}
+                value={modelSearchKeyword}
+                onChange={(value) => setModelSearchKeyword(value)}
+                style={{ marginBottom: '16px' }}
+                showClear
+              />
+              
+              <div style={{ 
+                display: 'grid', 
+                gridTemplateColumns: 'repeat(auto-fill, minmax(180px, 1fr))', 
+                gap: '10px' 
+              }}>
+                {currentTestChannel.models.split(',')
+                  .filter(model => model.toLowerCase().includes(modelSearchKeyword.toLowerCase()))
+                  .map((model, index) => {
+
+                    return (
+                      <Button
+                        key={index}
+                        theme="light"
+                        type="tertiary"
+                        style={{ 
+                          height: 'auto',
+                          padding: '8px 12px',
+                          textAlign: 'center',
+                        }}
+                        onClick={() => {
+                          testChannel(currentTestChannel, model);
+                        }}
+                      >
+                        {model}
+                      </Button>
+                    );
+                  })}
+              </div>
+              
+              {/* 显示搜索结果数量 */}
+              {modelSearchKeyword && (
+                <Typography.Text type="secondary" style={{ marginTop: '16px', display: 'block' }}>
+                  {t('找到')} {currentTestChannel.models.split(',').filter(model => 
+                    model.toLowerCase().includes(modelSearchKeyword.toLowerCase())
+                  ).length} {t('个模型')}
+                </Typography.Text>
+              )}
+            </div>
+          )}
+        </div>
+      </Modal>
     </>
   );
 };

web/src/components/HeaderBar.js

@@ -21,15 +21,17 @@ import {
   IconUser,
   IconLanguage
 } from '@douyinfe/semi-icons';
-import { Avatar, Button, Dropdown, Layout, Nav, Switch } from '@douyinfe/semi-ui';
+import { Avatar, Button, Dropdown, Layout, Nav, Switch, Tag } from '@douyinfe/semi-ui';
 import { stringToColor } from '../helpers/render';
 import Text from '@douyinfe/semi-ui/lib/es/typography/text';
 import { StyleContext } from '../context/Style/index.js';
+import { StatusContext } from '../context/Status/index.js';
 
 const HeaderBar = () => {
   const { t, i18n } = useTranslation();
   const [userState, userDispatch] = useContext(UserContext);
   const [styleState, styleDispatch] = useContext(StyleContext);
+  const [statusState, statusDispatch] = useContext(StatusContext);
   let navigate = useNavigate();
   const [currentLang, setCurrentLang] = useState(i18n.language);
 
@@ -40,6 +42,10 @@ const HeaderBar = () => {
   const isNewYear =
     (currentDate.getMonth() === 0 && currentDate.getDate() === 1);
 
+  // Check if self-use mode is enabled
+  const isSelfUseMode = statusState?.status?.self_use_mode_enabled || false;
+  const isDemoSiteMode = statusState?.status?.demo_site_enabled || false;
+
   let buttons = [
     {
       text: t('首页'),
@@ -166,7 +172,7 @@ const HeaderBar = () => {
             onSelect={(key) => {}}
             header={styleState.isMobile?{
               logo: (
-                <>
+                <div style={{ display: 'flex', alignItems: 'center', position: 'relative' }}>
                   {
                     !styleState.showSider ?
                       <Button icon={<IconMenu />} theme="light" aria-label={t('展开侧边栏')} onClick={
@@ -176,13 +182,52 @@ const HeaderBar = () => {
                         () => styleDispatch({ type: 'SET_SIDER', payload: false })
                       } />
                   }
-                </>
+                  {(isSelfUseMode || isDemoSiteMode) && (
+                    <Tag 
+                      color={isSelfUseMode ? 'purple' : 'blue'}
+                      style={{ 
+                        position: 'absolute',
+                        top: '-8px',
+                        right: '-15px',
+                        fontSize: '0.7rem',
+                        padding: '0 4px',
+                        height: 'auto',
+                        lineHeight: '1.2',
+                        zIndex: 1,
+                        pointerEvents: 'none'
+                      }}
+                    >
+                      {isSelfUseMode ? t('自用模式') : t('演示站点')}
+                    </Tag>
+                  )}
+                </div>
               ),
             }:{
               logo: (
                 <img src={logo} alt='logo' />
               ),
-              text: systemName,
+              text: (
+                <div style={{ position: 'relative', display: 'inline-block' }}>
+                  {systemName}
+                  {(isSelfUseMode || isDemoSiteMode) && (
+                    <Tag 
+                      color={isSelfUseMode ? 'purple' : 'blue'}
+                      style={{ 
+                        position: 'absolute', 
+                        top: '-10px', 
+                        right: '-25px', 
+                        fontSize: '0.7rem',
+                        padding: '0 4px',
+                        whiteSpace: 'nowrap',
+                        zIndex: 1,
+                        boxShadow: '0 0 3px rgba(255, 255, 255, 0.7)'
+                      }}
+                    >
+                      {isSelfUseMode ? t('自用模式') : t('演示站点')}
+                    </Tag>
+                  )}
+                </div>
+              ),
             }}
             items={buttons}
             footer={
@@ -266,7 +311,8 @@ const HeaderBar = () => {
                       icon={<IconUser />}
                     />
                     {
-                      !styleState.isMobile && (
+                      // Hide register option in self-use mode
+                      !styleState.isMobile && !isSelfUseMode && (
                         <Nav.Item
                           itemKey={'register'}
                           text={t('注册')}

web/src/components/LogsTable.js

@@ -464,6 +464,8 @@ const LogsTable = () => {
           other.model_ratio,
           other.model_price,
           other.group_ratio,
+          other.cache_tokens || 0,
+          other.cache_ratio || 1.0,
         );
         return (
             <Paragraph
@@ -665,6 +667,8 @@ const LogsTable = () => {
             other?.audio_ratio,
             other?.audio_completion_ratio,
             other.group_ratio,
+            other.cache_tokens || 0,
+            other.cache_ratio || 1.0,
           );
         } else {
           content = renderModelPrice(
@@ -674,6 +678,8 @@ const LogsTable = () => {
             other.model_price,
             other.completion_ratio,
             other.group_ratio,
+            other.cache_tokens || 0,
+            other.cache_ratio || 1.0,
           );
         }
         expandDataLocal.push({

web/src/components/OperationSetting.js

@@ -28,6 +28,7 @@ const OperationSetting = () => {
     PreConsumedQuota: 0,
     StreamCacheQueueLength: 0,
     ModelRatio: '',
+    CacheRatio: '',
     CompletionRatio: '',
     ModelPrice: '',
     GroupRatio: '',
@@ -60,6 +61,7 @@ const OperationSetting = () => {
     RetryTimes: 0,
     Chats: "[]",
     DemoSiteEnabled: false,
+    SelfUseModeEnabled: false,
     AutomaticDisableKeywords: '',
   });
 
@@ -76,7 +78,8 @@ const OperationSetting = () => {
           item.key === 'GroupRatio' ||
           item.key === 'UserUsableGroups' ||
           item.key === 'CompletionRatio' ||
-          item.key === 'ModelPrice'
+          item.key === 'ModelPrice' ||
+          item.key === 'CacheRatio'
         ) {
           item.value = JSON.stringify(JSON.parse(item.value), null, 2);
         }

web/src/components/OtherSetting.js

@@ -1,8 +1,10 @@
-import React, { useEffect, useRef, useState } from 'react';
-import { Banner, Button, Col, Form, Row } from '@douyinfe/semi-ui';
-import { API, showError, showSuccess } from '../helpers';
+import React, { useContext, useEffect, useRef, useState } from 'react';
+import { Banner, Button, Col, Form, Row, Modal, Space } from '@douyinfe/semi-ui';
+import { API, showError, showSuccess, timestamp2string } from '../helpers';
 import { marked } from 'marked';
 import { useTranslation } from 'react-i18next';
+import { StatusContext } from '../context/Status/index.js';
+import Text from '@douyinfe/semi-ui/lib/es/typography/text';
 
 const OtherSetting = () => {
   const { t } = useTranslation();
@@ -16,6 +18,7 @@ const OtherSetting = () => {
   });
   let [loading, setLoading] = useState(false);
   const [showUpdateModal, setShowUpdateModal] = useState(false);
+  const [statusState, statusDispatch] = useContext(StatusContext);
   const [updateData, setUpdateData] = useState({
     tag_name: '',
     content: '',
@@ -43,6 +46,7 @@ const OtherSetting = () => {
     HomePageContent: false,
     About: false,
     Footer: false,
+    CheckUpdate: false
   });
   const handleInputChange = async (value, e) => {
     const name = e.target.id;
@@ -145,23 +149,48 @@ const OtherSetting = () => {
     }
   };
 
-  const openGitHubRelease = () => {
-    window.location = 'https://github.com/songquanpeng/one-api/releases/latest';
-  };
-
   const checkUpdate = async () => {
-    const res = await API.get(
-      'https://api.github.com/repos/songquanpeng/one-api/releases/latest',
-    );
-    const { tag_name, body } = res.data;
-    if (tag_name === process.env.REACT_APP_VERSION) {
-      showSuccess(`已是最新版本：${tag_name}`);
-    } else {
-      setUpdateData({
-        tag_name: tag_name,
-        content: marked.parse(body),
-      });
-      setShowUpdateModal(true);
+    try {
+      setLoadingInput((loadingInput) => ({ ...loadingInput, CheckUpdate: true }));
+      // Use a CORS proxy to avoid direct cross-origin requests to GitHub API
+      // Option 1: Use a public CORS proxy service
+      // const proxyUrl = 'https://cors-anywhere.herokuapp.com/';
+      // const res = await API.get(
+      //   `${proxyUrl}https://api.github.com/repos/Calcium-Ion/new-api/releases/latest`,
+      // );
+      
+      // Option 2: Use the JSON proxy approach which often works better with GitHub API
+      const res = await fetch(
+        'https://api.github.com/repos/Calcium-Ion/new-api/releases/latest',
+        {
+          headers: {
+            'Accept': 'application/json',
+            'Content-Type': 'application/json',
+            // Adding User-Agent which is often required by GitHub API
+            'User-Agent': 'new-api-update-checker'
+          }
+        }
+      ).then(response => response.json());
+      
+      // Option 3: Use a local proxy endpoint
+      // Create a cached version of the response to avoid frequent GitHub API calls
+      // const res = await API.get('/api/status/github-latest-release');
+
+      const { tag_name, body } = res;
+      if (tag_name === statusState?.status?.version) {
+        showSuccess(`已是最新版本：${tag_name}`);
+      } else {
+        setUpdateData({
+          tag_name: tag_name,
+          content: marked.parse(body),
+        });
+        setShowUpdateModal(true);
+      }
+    } catch (error) {
+      console.error('Failed to check for updates:', error);
+      showError('检查更新失败，请稍后再试');
+    } finally {
+      setLoadingInput((loadingInput) => ({ ...loadingInput, CheckUpdate: false }));
     }
   };
   const getOptions = async () => {
@@ -186,9 +215,41 @@ const OtherSetting = () => {
     getOptions();
   }, []);
 
+  // Function to open GitHub release page
+  const openGitHubRelease = () => {
+    window.open(`https://github.com/Calcium-Ion/new-api/releases/tag/${updateData.tag_name}`, '_blank');
+  };
+
+  const getStartTimeString = () => {
+    const timestamp = statusState?.status?.start_time;
+    return statusState.status ? timestamp2string(timestamp) : '';
+  };
+
   return (
     <Row>
       <Col span={24}>
+        {/* 版本信息 */}
+        <Form style={{ marginBottom: 15 }}>
+          <Form.Section text={t('系统信息')}>
+            <Row>
+              <Col span={16}>
+                <Space>
+                  <Text>
+                    {t('当前版本')}：{statusState?.status?.version || t('未知')}
+                  </Text>
+                  <Button type="primary" onClick={checkUpdate} loading={loadingInput['CheckUpdate']}>
+                    {t('检查更新')}
+                  </Button>
+                </Space>
+              </Col>
+            </Row>
+            <Row>
+              <Col span={16}>
+                <Text>{t('启动时间')}：{getStartTimeString()}</Text>
+              </Col>
+            </Row>
+          </Form.Section>
+        </Form>
         {/* 通用设置 */}
         <Form
           values={inputs}
@@ -282,28 +343,25 @@ const OtherSetting = () => {
           </Form.Section>
         </Form>
       </Col>
-      {/*<Modal*/}
-      {/*  onClose={() => setShowUpdateModal(false)}*/}
-      {/*  onOpen={() => setShowUpdateModal(true)}*/}
-      {/*  open={showUpdateModal}*/}
-      {/*>*/}
-      {/*  <Modal.Header>新版本：{updateData.tag_name}</Modal.Header>*/}
-      {/*  <Modal.Content>*/}
-      {/*    <Modal.Description>*/}
-      {/*      <div dangerouslySetInnerHTML={{ __html: updateData.content }}></div>*/}
-      {/*    </Modal.Description>*/}
-      {/*  </Modal.Content>*/}
-      {/*  <Modal.Actions>*/}
-      {/*    <Button onClick={() => setShowUpdateModal(false)}>关闭</Button>*/}
-      {/*    <Button*/}
-      {/*      content='详情'*/}
-      {/*      onClick={() => {*/}
-      {/*        setShowUpdateModal(false);*/}
-      {/*        openGitHubRelease();*/}
-      {/*      }}*/}
-      {/*    />*/}
-      {/*  </Modal.Actions>*/}
-      {/*</Modal>*/}
+      <Modal
+        title={t('新版本') + '：' + updateData.tag_name}
+        visible={showUpdateModal}
+        onCancel={() => setShowUpdateModal(false)}
+        footer={[
+          <Button 
+            key="details" 
+            type="primary" 
+            onClick={() => {
+              setShowUpdateModal(false);
+              openGitHubRelease();
+            }}
+          >
+            {t('详情')}
+          </Button>
+        ]}
+      >
+        <div dangerouslySetInnerHTML={{ __html: updateData.content }}></div>
+      </Modal>
     </Row>
   );
 };

web/src/components/PersonalSetting.js

@@ -69,7 +69,11 @@ const PersonalSetting = () => {
     const [models, setModels] = useState([]);
     const [openTransfer, setOpenTransfer] = useState(false);
     const [transferAmount, setTransferAmount] = useState(0);
-    const [isModelsExpanded, setIsModelsExpanded] = useState(false);
+    const [isModelsExpanded, setIsModelsExpanded] = useState(() => {
+        // Initialize from localStorage if available
+        const savedState = localStorage.getItem('modelsExpanded');
+        return savedState ? JSON.parse(savedState) : false;
+    });
     const MODELS_DISPLAY_COUNT = 10;  // 默认显示的模型数量
     const [notificationSettings, setNotificationSettings] = useState({
         warningType: 'email',
@@ -124,6 +128,11 @@ const PersonalSetting = () => {
         }
     }, [userState?.user?.setting]);
 
+    // Save models expanded state to localStorage whenever it changes
+    useEffect(() => {
+        localStorage.setItem('modelsExpanded', JSON.stringify(isModelsExpanded));
+    }, [isModelsExpanded]);
+
     const handleInputChange = (name, value) => {
         setInputs((inputs) => ({...inputs, [name]: value}));
     };
@@ -384,7 +393,7 @@ const PersonalSetting = () => {
                             </div>
                         </div>
                     </Modal>
-                    <div style={{marginTop: 20}}>
+                    <div>
                         <Card
                             title={
                                 <Card.Meta

web/src/components/SiderBar.js

@@ -28,7 +28,7 @@ import {
   IconSetting,
   IconUser
 } from '@douyinfe/semi-icons';
-import { Avatar, Dropdown, Layout, Nav, Switch } from '@douyinfe/semi-ui';
+import { Avatar, Dropdown, Layout, Nav, Switch, Divider } from '@douyinfe/semi-ui';
 import { setStatusData } from '../helpers/data.js';
 import { stringToColor } from '../helpers/render.js';
 import { useSetTheme, useTheme } from '../context/Theme/index.js';
@@ -65,35 +65,11 @@ const SiderBar = () => {
     pricing: '/pricing',
     task: '/task',
     playground: '/playground',
+    personal: '/personal',
   };
 
-  const headerButtons = useMemo(
+  const workspaceItems = useMemo(
     () => [
-      {
-        text: 'Playground',
-        itemKey: 'playground',
-        to: '/playground',
-        icon: <IconCommentStroked />,
-      },
-      {
-        text: t('渠道'),
-        itemKey: 'channel',
-        to: '/channel',
-        icon: <IconLayers />,
-        className: isAdmin() ? '' : 'tableHiddle',
-      },
-      {
-        text: t('聊天'),
-        itemKey: 'chat',
-        items: chatItems,
-        icon: <IconComment />,
-      },
-      {
-        text: t('令牌'),
-        itemKey: 'token',
-        to: '/token',
-        icon: <IconKey />,
-      },
       {
         text: t('数据看板'),
         itemKey: 'detail',
@@ -105,33 +81,19 @@ const SiderBar = () => {
             : 'tableHiddle',
       },
       {
-        text: t('兑换码'),
-        itemKey: 'redemption',
-        to: '/redemption',
-        icon: <IconGift />,
-        className: isAdmin() ? '' : 'tableHiddle',
+        text: t('API令牌'),
+        itemKey: 'token',
+        to: '/token',
+        icon: <IconKey />,
       },
       {
-        text: t('钱包'),
-        itemKey: 'topup',
-        to: '/topup',
-        icon: <IconCreditCard />,
-      },
-      {
-        text: t('用户管理'),
-        itemKey: 'user',
-        to: '/user',
-        icon: <IconUser />,
-        className: isAdmin() ? '' : 'tableHiddle',
-      },
-      {
-        text: t('日志'),
+        text: t('使用日志'),
         itemKey: 'log',
         to: '/log',
         icon: <IconHistogram />,
       },
       {
-        text: t('绘图'),
+        text: t('绘图日志'),
         itemKey: 'midjourney',
         to: '/midjourney',
         icon: <IconImage />,
@@ -141,75 +103,148 @@ const SiderBar = () => {
             : 'tableHiddle',
       },
       {
-        text: t('异步任务'),
+        text: t('任务日志'),
         itemKey: 'task',
         to: '/task',
         icon: <IconChecklistStroked />,
         className:
-            localStorage.getItem('enable_task') === 'true'
-                ? ''
-                : 'tableHiddle',
-      },
-      {
-        text: t('设置'),
-        itemKey: 'setting',
-        to: '/setting',
-        icon: <IconSetting />,
-      },
+          localStorage.getItem('enable_task') === 'true'
+            ? ''
+            : 'tableHiddle',
+      }
     ],
     [
       localStorage.getItem('enable_data_export'),
       localStorage.getItem('enable_drawing'),
       localStorage.getItem('enable_task'),
-      localStorage.getItem('chat_link'),
-      chatItems,
-      isAdmin(),
       t,
     ],
   );
 
+  const financeItems = useMemo(
+    () => [
+      {
+        text: t('钱包'),
+        itemKey: 'topup',
+        to: '/topup',
+        icon: <IconCreditCard />,
+      },
+      {
+        text: t('个人设置'),
+        itemKey: 'personal',
+        to: '/personal',
+        icon: <IconUser />,
+      },
+    ],
+    [t],
+  );
+
+  const adminItems = useMemo(
+    () => [
+      {
+        text: t('渠道'),
+        itemKey: 'channel',
+        to: '/channel',
+        icon: <IconLayers />,
+        className: isAdmin() ? '' : 'tableHiddle',
+      },
+      {
+        text: t('兑换码'),
+        itemKey: 'redemption',
+        to: '/redemption',
+        icon: <IconGift />,
+        className: isAdmin() ? '' : 'tableHiddle',
+      },
+      {
+        text: t('用户管理'),
+        itemKey: 'user',
+        to: '/user',
+        icon: <IconUser />,
+      },
+      {
+        text: t('系统设置'),
+        itemKey: 'setting',
+        to: '/setting',
+        icon: <IconSetting />,
+      },
+    ],
+    [isAdmin(), t],
+  );
+
+  const chatMenuItems = useMemo(
+    () => [
+      {
+        text: 'Playground',
+        itemKey: 'playground',
+        to: '/playground',
+        icon: <IconCommentStroked />,
+      },
+      {
+        text: t('聊天'),
+        itemKey: 'chat',
+        items: chatItems,
+        icon: <IconComment />,
+      },
+    ],
+    [chatItems, t],
+  );
+
   useEffect(() => {
     let localKey = window.location.pathname.split('/')[1];
     if (localKey === '') {
       localKey = 'home';
     }
     setSelectedKeys([localKey]);
-    
+
     let chatLink = localStorage.getItem('chat_link');
     if (!chatLink) {
-        let chats = localStorage.getItem('chats');
-        if (chats) {
-            // console.log(chats);
-            try {
-                chats = JSON.parse(chats);
-                if (Array.isArray(chats)) {
-                    let chatItems = [];
-                    for (let i = 0; i < chats.length; i++) {
-                        let chat = {};
-                        for (let key in chats[i]) {
-                            chat.text = key;
-                            chat.itemKey = 'chat' + i;
-                            chat.to = '/chat/' + i;
-                        }
-                        // setRouterMap({ ...routerMap, chat: '/chat/' + i })
-                        chatItems.push(chat);
-                    }
-                    setChatItems(chatItems);
-                }
-            } catch (e) {
-                console.error(e);
-                showError('聊天数据解析失败')
+      let chats = localStorage.getItem('chats');
+      if (chats) {
+        // console.log(chats);
+        try {
+          chats = JSON.parse(chats);
+          if (Array.isArray(chats)) {
+            let chatItems = [];
+            for (let i = 0; i < chats.length; i++) {
+              let chat = {};
+              for (let key in chats[i]) {
+                chat.text = key;
+                chat.itemKey = 'chat' + i;
+                chat.to = '/chat/' + i;
+              }
+              // setRouterMap({ ...routerMap, chat: '/chat/' + i })
+              chatItems.push(chat);
             }
+            setChatItems(chatItems);
+          }
+        } catch (e) {
+          console.error(e);
+          showError('聊天数据解析失败')
         }
+      }
     }
-    
+
     setIsCollapsed(localStorage.getItem('default_collapse_sidebar') === 'true');
   }, []);
 
+  // Custom divider style
+  const dividerStyle = {
+    margin: '8px 0',
+    opacity: 0.6,
+  };
+
+  // Custom group label style
+  const groupLabelStyle = {
+    padding: '8px 16px',
+    color: 'var(--semi-color-text-2)',
+    fontSize: '12px',
+    fontWeight: 'normal',
+  };
+
   return (
     <>
       <Nav
-        style={{ maxWidth: 220, height: '100%' }}
+        style={{ maxWidth: 200, height: '100%' }}
         defaultIsCollapsed={
           localStorage.getItem('default_collapse_sidebar') === 'true'
         }
@@ -219,27 +254,27 @@ const SiderBar = () => {
         }}
         selectedKeys={selectedKeys}
         renderWrapper={({ itemElement, isSubNav, isInSubNav, props }) => {
-            let chatLink = localStorage.getItem('chat_link');
-            if (!chatLink) {
-                let chats = localStorage.getItem('chats');
-                if (chats) {
-                    chats = JSON.parse(chats);
-                    if (Array.isArray(chats) && chats.length > 0) {
-                        for (let i = 0; i < chats.length; i++) {
-                            routerMap['chat' + i] = '/chat/' + i;
-                        }
-                        if (chats.length > 1) {
-                            // delete /chat
-                            if (routerMap['chat']) {
-                                delete routerMap['chat'];
-                            }
-                        } else {
-                            // rename /chat to /chat/0
-                            routerMap['chat'] = '/chat/0';
-                        }
-                    }
+          let chatLink = localStorage.getItem('chat_link');
+          if (!chatLink) {
+            let chats = localStorage.getItem('chats');
+            if (chats) {
+              chats = JSON.parse(chats);
+              if (Array.isArray(chats) && chats.length > 0) {
+                for (let i = 0; i < chats.length; i++) {
+                  routerMap['chat' + i] = '/chat/' + i;
                 }
+                if (chats.length > 1) {
+                  // delete /chat
+                  if (routerMap['chat']) {
+                    delete routerMap['chat'];
+                  }
+                } else {
+                  // rename /chat to /chat/0
+                  routerMap['chat'] = '/chat/0';
+                }
+              }
             }
+          }
           return (
             <Link
               style={{ textDecoration: 'none' }}
@@ -249,7 +284,6 @@ const SiderBar = () => {
             </Link>
           );
         }}
-        items={headerButtons}
         onSelect={(key) => {
           if (key.itemKey.toString().startsWith('chat')) {
             styleDispatch({ type: 'SET_INNER_PADDING', payload: false });
@@ -258,12 +292,101 @@ const SiderBar = () => {
           }
           setSelectedKeys([key.itemKey]);
         }}
-        footer={
-          <>
-          </>
-        }
       >
-        <Nav.Footer collapseButton={true}></Nav.Footer>
+        {/* Chat Section - Only show if there are chat items */}
+        {chatItems.length > 0 && (
+          <>
+            {chatMenuItems.map((item) => {
+              if (item.items && item.items.length > 0) {
+                return (
+                  <Nav.Sub
+                    key={item.itemKey}
+                    itemKey={item.itemKey}
+                    text={item.text}
+                    icon={item.icon}
+                  >
+                    {item.items.map((subItem) => (
+                      <Nav.Item
+                        key={subItem.itemKey}
+                        itemKey={subItem.itemKey}
+                        text={subItem.text}
+                      />
+                    ))}
+                  </Nav.Sub>
+                );
+              } else {
+                return (
+                  <Nav.Item
+                    key={item.itemKey}
+                    itemKey={item.itemKey}
+                    text={item.text}
+                    icon={item.icon}
+                  />
+                );
+              }
+            })}
+          </>
+        )}
+
+        {/* Divider */}
+        <Divider style={dividerStyle} />
+
+        {/* Workspace Section */}
+        {!isCollapsed && <div style={groupLabelStyle}>{t('控制台')}</div>}
+        {workspaceItems.map((item) => (
+          <Nav.Item
+            key={item.itemKey}
+            itemKey={item.itemKey}
+            text={item.text}
+            icon={item.icon}
+            className={item.className}
+          />
+        ))}
+
+        {/* Divider */}
+        <Divider style={dividerStyle} />
+
+        {/* Finance Management Section */}
+        {!isCollapsed && <div style={groupLabelStyle}>{t('个人中心')}</div>}
+        {financeItems.map((item) => (
+          <Nav.Item
+            key={item.itemKey}
+            itemKey={item.itemKey}
+            text={item.text}
+            icon={item.icon}
+            className={item.className}
+          />
+        ))}
+
+        {isAdmin() && (
+          <>
+            {/* Divider */}
+            <Divider style={dividerStyle} />
+
+            {/* Admin Section */}
+            {adminItems.map((item) => (
+              <Nav.Item
+                key={item.itemKey}
+                itemKey={item.itemKey}
+                text={item.text}
+                icon={item.icon}
+                className={item.className}
+              />
+            ))}
+          </>
+        )}
+
+        <Nav.Footer
+          collapseButton={true}
+          collapseText={(collapsed)=>
+            {
+              if(collapsed){
+                return t('展开侧边栏')
+              }
+                return t('收起侧边栏')
+            }
+          }
+        />
       </Nav>
     </>
   );

web/src/constants/channel.constants.js

@@ -82,7 +82,7 @@ export const CHANNEL_OPTIONS = [
   {
     value: 45,
     color: 'blue',
-    label: '火山方舟（豆包）'
+    label: '字节火山方舟、豆包、DeepSeek通用'
   },
   { value: 25, color: 'green', label: 'Moonshot' },
   { value: 19, color: 'blue', label: '360 智脑' },

web/src/helpers/render.js

@@ -298,6 +298,8 @@ export function renderModelPrice(
   modelPrice = -1,
   completionRatio,
   groupRatio,
+  cacheTokens = 0,
+  cacheRatio = 1.0,
 ) {
   if (modelPrice !== -1) {
     return i18next.t('模型价格：${{price}} * 分组倍率：{{ratio}} = ${{total}}', {
@@ -311,9 +313,15 @@ export function renderModelPrice(
     }
     let inputRatioPrice = modelRatio * 2.0;
     let completionRatioPrice = modelRatio * 2.0 * completionRatio;
+    let cacheRatioPrice = modelRatio * 2.0 * cacheRatio;
+    
+    // Calculate effective input tokens (non-cached + cached with ratio applied)
+    const effectiveInputTokens = (inputTokens - cacheTokens) + (cacheTokens * cacheRatio);
+    
     let price =
-      (inputTokens / 1000000) * inputRatioPrice * groupRatio +
+      (effectiveInputTokens / 1000000) * inputRatioPrice * groupRatio +
       (completionTokens / 1000000) * completionRatioPrice * groupRatio;
+    
     return (
       <>
         <article>
@@ -327,16 +335,36 @@ export function renderModelPrice(
             ratio: groupRatio,
             total: completionRatioPrice * groupRatio
           })}</p>
+          {cacheTokens > 0 && (
+            <p>{i18next.t('缓存：${{price}} * {{ratio}} = ${{total}} / 1M tokens (缓存比例: {{cacheRatio}})', {
+              price: cacheRatioPrice,
+              ratio: groupRatio,
+              total: cacheRatioPrice * groupRatio,
+              cacheRatio: cacheRatio
+            })}</p>
+          )}
           <p></p>
           <p>
-            {i18next.t('提示 {{input}} tokens / 1M tokens * ${{price}} + 补全 {{completion}} tokens / 1M tokens * ${{compPrice}} * 分组 {{ratio}} = ${{total}}', {
-              input: inputTokens,
-              price: inputRatioPrice,
-              completion: completionTokens,
-              compPrice: completionRatioPrice,
-              ratio: groupRatio,
-              total: price.toFixed(6)
-            })}
+            {cacheTokens > 0 ? 
+              i18next.t('提示 {{nonCacheInput}} tokens + 缓存 {{cacheInput}} tokens * {{cacheRatio}} / 1M tokens * ${{price}} + 补全 {{completion}} tokens / 1M tokens * ${{compPrice}} * 分组 {{ratio}} = ${{total}}', {
+                nonCacheInput: inputTokens - cacheTokens,
+                cacheInput: cacheTokens,
+                cacheRatio: cacheRatio,
+                price: inputRatioPrice,
+                completion: completionTokens,
+                compPrice: completionRatioPrice,
+                ratio: groupRatio,
+                total: price.toFixed(6)
+              }) :
+              i18next.t('提示 {{input}} tokens / 1M tokens * ${{price}} + 补全 {{completion}} tokens / 1M tokens * ${{compPrice}} * 分组 {{ratio}} = ${{total}}', {
+                input: inputTokens,
+                price: inputRatioPrice,
+                completion: completionTokens,
+                compPrice: completionRatioPrice,
+                ratio: groupRatio,
+                total: price.toFixed(6)
+              })
+            }
           </p>
           <p>{i18next.t('仅供参考，以实际扣费为准')}</p>
         </article>
@@ -349,6 +377,8 @@ export function renderModelPriceSimple(
   modelRatio,
   modelPrice = -1,
   groupRatio,
+  cacheTokens = 0,
+  cacheRatio = 1.0,
 ) {
   if (modelPrice !== -1) {
     return i18next.t('价格：${{price}} * 分组：{{ratio}}', {
@@ -356,10 +386,18 @@ export function renderModelPriceSimple(
       ratio: groupRatio
     });
   } else {
-    return i18next.t('模型: {{ratio}} * 分组: {{groupRatio}}', {
-      ratio: modelRatio,
-      groupRatio: groupRatio
-    });
+    if (cacheTokens !== 0) {
+      return i18next.t('模型: {{ratio}} * 分组: {{groupRatio}} * 缓存比例: {{cacheRatio}}', {
+        ratio: modelRatio,
+        groupRatio: groupRatio,
+        cacheRatio: cacheRatio
+      });
+    } else {
+      return i18next.t('模型: {{ratio}} * 分组: {{groupRatio}}', {
+        ratio: modelRatio,
+        groupRatio: groupRatio
+      });
+    }
   }
 }
 
@@ -374,6 +412,8 @@ export function renderAudioModelPrice(
   audioRatio,
   audioCompletionRatio,
   groupRatio,
+  cacheTokens = 0,
+  cacheRatio = 1.0,
 ) {
   // 1 ratio = $0.002 / 1K tokens
   if (modelPrice !== -1) {
@@ -388,8 +428,13 @@ export function renderAudioModelPrice(
     // 这里的 *2 是因为 1倍率=0.002刀，请勿删除
     let inputRatioPrice = modelRatio * 2.0;
     let completionRatioPrice = modelRatio * 2.0 * completionRatio;
+    let cacheRatioPrice = modelRatio * 2.0 * cacheRatio;
+    
+    // Calculate effective input tokens (non-cached + cached with ratio applied)
+    const effectiveInputTokens = (inputTokens - cacheTokens) + (cacheTokens * cacheRatio);
+    
     let price =
-      (inputTokens / 1000000) * inputRatioPrice * groupRatio +
+      (effectiveInputTokens / 1000000) * inputRatioPrice * groupRatio +
       (completionTokens / 1000000) * completionRatioPrice * groupRatio +
       (audioInputTokens / 1000000) * inputRatioPrice * audioRatio * groupRatio +
       (audioCompletionTokens / 1000000) * inputRatioPrice * audioRatio * audioCompletionRatio * groupRatio;
@@ -406,6 +451,14 @@ export function renderAudioModelPrice(
             ratio: groupRatio,
             total: completionRatioPrice * groupRatio
           })}</p>
+          {cacheTokens > 0 && (
+            <p>{i18next.t('缓存：${{price}} * {{ratio}} = ${{total}} / 1M tokens (缓存比例: {{cacheRatio}})', {
+              price: cacheRatioPrice,
+              ratio: groupRatio,
+              total: cacheRatioPrice * groupRatio,
+              cacheRatio: cacheRatio
+            })}</p>
+          )}
           <p>{i18next.t('音频提示：${{price}} * {{ratio}} * {{audioRatio}} = ${{total}} / 1M tokens', {
             price: inputRatioPrice,
             ratio: groupRatio,
@@ -420,12 +473,22 @@ export function renderAudioModelPrice(
             total: inputRatioPrice * audioRatio * audioCompletionRatio * groupRatio
           })}</p>
           <p>
-            {i18next.t('文字提示 {{input}} tokens / 1M tokens * ${{price}} + 文字补全 {{completion}} tokens / 1M tokens * ${{compPrice}} +', {
-              input: inputTokens,
-              price: inputRatioPrice,
-              completion: completionTokens,
-              compPrice: completionRatioPrice
-            })}
+            {cacheTokens > 0 ? 
+              i18next.t('文字提示 {{nonCacheInput}} tokens + 文字缓存 {{cacheInput}} tokens * {{cacheRatio}} / 1M tokens * ${{price}} + 文字补全 {{completion}} tokens / 1M tokens * ${{compPrice}} +', {
+                nonCacheInput: inputTokens - cacheTokens,
+                cacheInput: cacheTokens,
+                cacheRatio: cacheRatio,
+                price: inputRatioPrice,
+                completion: completionTokens,
+                compPrice: completionRatioPrice
+              }) :
+              i18next.t('文字提示 {{input}} tokens / 1M tokens * ${{price}} + 文字补全 {{completion}} tokens / 1M tokens * ${{compPrice}} +', {
+                input: inputTokens,
+                price: inputRatioPrice,
+                completion: completionTokens,
+                compPrice: completionRatioPrice
+              })
+            }
           </p>
           <p>
             {i18next.t('音频提示 {{input}} tokens / 1M tokens * ${{price}} * {{audioRatio}} + 音频补全 {{completion}} tokens / 1M tokens * ${{price}} * {{audioRatio}} * {{audioCompRatio}}', {

web/src/i18n/locales/en.json

@@ -621,6 +621,7 @@
   "窗口等待": "window wait",
   "失败": "Failed",
   "绘图": "Drawing",
+  "绘图日志": "Drawing log",
   "放大": "Upscalers",
   "微妙放大": "Upscale (Subtle)",
   "创造放大": "Upscale (Creative)",
@@ -1120,7 +1121,7 @@
   "知识库 ID": "Knowledge Base ID",
   "请输入知识库 ID，例如：123456": "Please enter knowledge base ID, e.g.: 123456",
   "可选值": "Optional value",
-  "异步任务": "Async task",
+  "任务日志": "Task log",
   "你好": "Hello",
   "你好，请问有什么可以帮助您的吗？": "Hello, how may I help you?",
   "用户分组": "Your default group",
@@ -1317,5 +1318,25 @@
   "当前设置类型: ": "Current setting type: ",
   "固定价格值": "Fixed Price Value",
   "未设置倍率模型": "Models without ratio settings",
-  "模型倍率和补全倍率同时设置": "Both model ratio and completion ratio are set"
+  "模型倍率和补全倍率同时设置": "Both model ratio and completion ratio are set",
+  "自用模式": "Self-use mode",
+  "开启后不限制：必须设置模型倍率": "After enabling, no limit: must set model ratio",
+  "演示站点模式": "Demo site mode",
+  "当前版本": "Current version",
+  "Gemini设置": "Gemini settings",
+  "Gemini安全设置": "Gemini safety settings",
+  "default为默认设置，可单独设置每个分类的安全等级": "\"default\" is the default setting, and each category can be set separately",
+  "Gemini版本设置": "Gemini version settings",
+  "default为默认设置，可单独设置每个模型的版本": "\"default\" is the default setting, and each model can be set separately",
+  "Claude设置": "Claude settings",
+  "Claude请求头覆盖": "Claude request header override",
+  "示例": "Example",
+  "缺省 MaxTokens": "Default MaxTokens",
+  "启用Claude思考适配（-thinking后缀）": "Enable Claude thinking adaptation (-thinking suffix)",
+  "Claude思考适配 BudgetTokens = MaxTokens * BudgetTokens 百分比": "Claude thinking adaptation BudgetTokens = MaxTokens * BudgetTokens percentage",
+  "思考适配 BudgetTokens 百分比": "Thinking adaptation BudgetTokens percentage",
+  "0.1-1之间的小数": "Decimal between 0.1 and 1",
+  "模型相关设置": "Model related settings",
+  "收起侧边栏": "Collapse sidebar",
+  "展开侧边栏": "Expand sidebar"
 }

web/src/pages/Setting/Model/SettingClaudeModel.js

@@ -18,6 +18,8 @@ const CLAUDE_HEADER = {
 
 const CLAUDE_DEFAULT_MAX_TOKENS = {
   'default': 8192,
+  "claude-3-haiku-20240307": 4096,
+  "claude-3-opus-20240229": 4096,
   'claude-3-7-sonnet-20250219-thinking': 8192,
 }
 

web/src/pages/Setting/Operation/ModelRatioSettings.js

@@ -15,6 +15,7 @@ export default function ModelRatioSettings(props) {
   const [inputs, setInputs] = useState({
     ModelPrice: '',
     ModelRatio: '',
+    CacheRatio: '',
     CompletionRatio: '',
   });
   const refForm = useRef();
@@ -139,6 +140,25 @@ export default function ModelRatioSettings(props) {
               />
             </Col>
           </Row>
+          <Row gutter={16}>
+            <Col span={16}>
+              <Form.TextArea
+                label={t('提示缓存倍率')}
+                placeholder={t('为一个 JSON 文本，键为模型名称，值为倍率')}
+                field={'CacheRatio'}
+                autosize={{ minRows: 6, maxRows: 12 }}
+                trigger='blur'
+                stopValidateWithError
+                rules={[
+                  {
+                    validator: (rule, value) => verifyJSON(value),
+                    message: '不是合法的 JSON 字符串'
+                  }
+                ]}
+                onChange={(value) => setInputs({ ...inputs, CacheRatio: value })}
+              />
+            </Col>
+          </Row>
           <Row gutter={16}>
             <Col span={16}>
               <Form.TextArea

web/src/pages/Setting/Operation/ModelRationNotSetEditor.js

@@ -55,18 +55,17 @@ export default function ModelRatioNotSetEditor(props) {
       const unsetModels = enabledModels.filter(modelName => {
         const hasPrice = modelPrice[modelName] !== undefined;
         const hasRatio = modelRatio[modelName] !== undefined;
-        const hasCompletionRatio = completionRatio[modelName] !== undefined;
         
-        // 如果模型既没有价格也没有倍率设置，则显示
-        return !(hasPrice || (hasRatio && hasCompletionRatio));
+        // 如果模型没有价格或者没有倍率设置，则显示
+        return !hasPrice && !hasRatio;
       });
 
       // 创建模型数据
       const modelData = unsetModels.map(name => ({
         name,
-        price: '',
-        ratio: '',
-        completionRatio: ''
+        price: modelPrice[name] || '',
+        ratio: modelRatio[name] || '',
+        completionRatio: completionRatio[name] || ''
       }));
 
       setModels(modelData);

web/src/pages/Setting/Operation/SettingsGeneral.js

@@ -22,6 +22,7 @@ export default function GeneralSettings(props) {
     DisplayTokenStatEnabled: false,
     DefaultCollapseSidebar: false,
     DemoSiteEnabled: false,
+    SelfUseModeEnabled: false,
   });
   const refForm = useRef();
   const [inputsRow, setInputsRow] = useState(inputs);
@@ -205,6 +206,22 @@ export default function GeneralSettings(props) {
                   }
                 />
               </Col>
+              <Col span={8}>
+                <Form.Switch
+                  field={'SelfUseModeEnabled'}
+                  label={t('自用模式')}
+                  extraText={t('开启后不限制：必须设置模型倍率')}
+                  size='default'
+                  checkedText='｜'
+                  uncheckedText='〇'
+                  onChange={(value) =>
+                    setInputs({
+                      ...inputs,
+                      SelfUseModeEnabled: value
+                    })
+                  }
+                />
+              </Col>
             </Row>
             <Row>
               <Button size='default' onClick={onSubmit}>

web/src/pages/Setting/index.js

@@ -18,9 +18,6 @@ const Setting = () => {
   const [tabActiveKey, setTabActiveKey] = useState('1');
   let panes = [
     {
-      tab: t('个人设置'),
-      content: <PersonalSetting />,
-      itemKey: 'personal',
     },
   ];
 
@@ -61,7 +58,7 @@ const Setting = () => {
     if (tab) {
       setTabActiveKey(tab);
     } else {
-      onChangeTab('personal');
+      onChangeTab('operation');
     }
   }, [location.search]);
   return (