feat: support tencent v3 api (#1542)

Co-authored-by: lihangfu <hfli8@iflytek.com>

.github/workflows/docker-image-amd64-en.yml

@@ -3,7 +3,7 @@ name: Publish Docker image (amd64, English)
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
   workflow_dispatch:
     inputs:
       name:

.github/workflows/docker-image-amd64.yml

@@ -3,7 +3,7 @@ name: Publish Docker image (amd64)
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
   workflow_dispatch:
     inputs:
       name:

.github/workflows/docker-image-arm64.yml

@@ -3,7 +3,7 @@ name: Publish Docker image (arm64)
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
       - '!*-alpha*'
   workflow_dispatch:
     inputs:

.github/workflows/linux-release.yml

@@ -5,7 +5,7 @@ permissions:
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
       - '!*-alpha*'
   workflow_dispatch:
     inputs:

.github/workflows/macos-release.yml

@@ -5,7 +5,7 @@ permissions:
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
       - '!*-alpha*'
   workflow_dispatch:
     inputs:

.github/workflows/windows-release.yml

@@ -5,7 +5,7 @@ permissions:
 on:
   push:
     tags:
-      - '*'
+      - 'v*.*.*'
       - '!*-alpha*'
   workflow_dispatch:
     inputs:

README.md

@@ -68,6 +68,7 @@ _✨ 通过标准的 OpenAI API 格式访问所有的大模型，开箱即用
    + [x] [Anthropic Claude 系列模型](https://anthropic.com) (支持 AWS Claude)
    + [x] [Google PaLM2/Gemini 系列模型](https://developers.generativeai.google)
    + [x] [Mistral 系列模型](https://mistral.ai/)
+   + [x] [字节跳动豆包大模型](https://console.volcengine.com/ark/region:ark+cn-beijing/model)
    + [x] [百度文心一言系列模型](https://cloud.baidu.com/doc/WENXINWORKSHOP/index.html)
    + [x] [阿里通义千问系列模型](https://help.aliyun.com/document_detail/2400395.html)
    + [x] [讯飞星火认知大模型](https://www.xfyun.cn/doc/spark/Web.html)
@@ -76,12 +77,17 @@ _✨ 通过标准的 OpenAI API 格式访问所有的大模型，开箱即用
    + [x] [腾讯混元大模型](https://cloud.tencent.com/document/product/1729)
    + [x] [Moonshot AI](https://platform.moonshot.cn/)
    + [x] [百川大模型](https://platform.baichuan-ai.com)
-   + [ ] [字节云雀大模型](https://www.volcengine.com/product/ark) (WIP)
    + [x] [MINIMAX](https://api.minimax.chat/)
    + [x] [Groq](https://wow.groq.com/)
    + [x] [Ollama](https://github.com/ollama/ollama)
    + [x] [零一万物](https://platform.lingyiwanwu.com/)
    + [x] [阶跃星辰](https://platform.stepfun.com/)
+   + [x] [Coze](https://www.coze.com/)
+   + [x] [Cohere](https://cohere.com/)
+   + [x] [DeepSeek](https://www.deepseek.com/)
+   + [x] [Cloudflare Workers AI](https://developers.cloudflare.com/workers-ai/)
+   + [x] [DeepL](https://www.deepl.com/)
+   + [x] [together.ai](https://www.together.ai/)
 2. 支持配置镜像以及众多[第三方代理服务](https://iamazing.cn/page/openai-api-third-party-services)。
 3. 支持通过**负载均衡**的方式访问多个渠道。
 4. 支持 **stream 模式**，可以通过流式传输实现打字机效果。
@@ -378,14 +384,17 @@ graph LR
     + `TIKTOKEN_CACHE_DIR`：默认程序启动时会联网下载一些通用的词元的编码，如：`gpt-3.5-turbo`，在一些网络环境不稳定，或者离线情况，可能会导致启动有问题，可以配置此目录缓存数据，可迁移到离线环境。
     + `DATA_GYM_CACHE_DIR`：目前该配置作用与 `TIKTOKEN_CACHE_DIR` 一致，但是优先级没有它高。
 17. `RELAY_TIMEOUT`：中继超时设置，单位为秒，默认不设置超时时间。
-18. `SQLITE_BUSY_TIMEOUT`：SQLite 锁等待超时设置，单位为毫秒，默认 `3000`。
-19. `GEMINI_SAFETY_SETTING`：Gemini 的安全设置，默认 `BLOCK_NONE`。
-20. `GEMINI_VERSION`：One API 所使用的 Gemini 版本，默认为 `v1`。
-21. `THEME`：系统的主题设置，默认为 `default`，具体可选值参考[此处](./web/README.md)。
-22. `ENABLE_METRIC`：是否根据请求成功率禁用渠道，默认不开启，可选值为 `true` 和 `false`。
-23. `METRIC_QUEUE_SIZE`：请求成功率统计队列大小，默认为 `10`。
-24. `METRIC_SUCCESS_RATE_THRESHOLD`：请求成功率阈值，默认为 `0.8`。
-25. `INITIAL_ROOT_TOKEN`：如果设置了该值，则在系统首次启动时会自动创建一个值为该环境变量值的 root 用户令牌。
+18. `RELAY_PROXY`：设置后使用该代理来请求 API。
+19. `USER_CONTENT_REQUEST_TIMEOUT`：用户上传内容下载超时时间，单位为秒。
+20. `USER_CONTENT_REQUEST_PROXY`：设置后使用该代理来请求用户上传的内容，例如图片。
+21. `SQLITE_BUSY_TIMEOUT`：SQLite 锁等待超时设置，单位为毫秒，默认 `3000`。
+22. `GEMINI_SAFETY_SETTING`：Gemini 的安全设置，默认 `BLOCK_NONE`。
+23. `GEMINI_VERSION`：One API 所使用的 Gemini 版本，默认为 `v1`。
+24. `THEME`：系统的主题设置，默认为 `default`，具体可选值参考[此处](./web/README.md)。
+25. `ENABLE_METRIC`：是否根据请求成功率禁用渠道，默认不开启，可选值为 `true` 和 `false`。
+26. `METRIC_QUEUE_SIZE`：请求成功率统计队列大小，默认为 `10`。
+27. `METRIC_SUCCESS_RATE_THRESHOLD`：请求成功率阈值，默认为 `0.8`。
+28. `INITIAL_ROOT_TOKEN`：如果设置了该值，则在系统首次启动时会自动创建一个值为该环境变量值的 root 用户令牌。
 
 ### 命令行参数
 1. `--port <port_number>`: 指定服务器监听的端口号，默认为 `3000`。

common/client/init.go

@@ -0,0 +1,60 @@
+package client
+
+import (
+	"fmt"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"net/http"
+	"net/url"
+	"time"
+)
+
+var HTTPClient *http.Client
+var ImpatientHTTPClient *http.Client
+var UserContentRequestHTTPClient *http.Client
+
+func Init() {
+	if config.UserContentRequestProxy != "" {
+		logger.SysLog(fmt.Sprintf("using %s as proxy to fetch user content", config.UserContentRequestProxy))
+		proxyURL, err := url.Parse(config.UserContentRequestProxy)
+		if err != nil {
+			logger.FatalLog(fmt.Sprintf("USER_CONTENT_REQUEST_PROXY set but invalid: %s", config.UserContentRequestProxy))
+		}
+		transport := &http.Transport{
+			Proxy: http.ProxyURL(proxyURL),
+		}
+		UserContentRequestHTTPClient = &http.Client{
+			Transport: transport,
+			Timeout:   time.Second * time.Duration(config.UserContentRequestTimeout),
+		}
+	} else {
+		UserContentRequestHTTPClient = &http.Client{}
+	}
+	var transport http.RoundTripper
+	if config.RelayProxy != "" {
+		logger.SysLog(fmt.Sprintf("using %s as api relay proxy", config.RelayProxy))
+		proxyURL, err := url.Parse(config.RelayProxy)
+		if err != nil {
+			logger.FatalLog(fmt.Sprintf("USER_CONTENT_REQUEST_PROXY set but invalid: %s", config.UserContentRequestProxy))
+		}
+		transport = &http.Transport{
+			Proxy: http.ProxyURL(proxyURL),
+		}
+	}
+
+	if config.RelayTimeout == 0 {
+		HTTPClient = &http.Client{
+			Transport: transport,
+		}
+	} else {
+		HTTPClient = &http.Client{
+			Timeout:   time.Duration(config.RelayTimeout) * time.Second,
+			Transport: transport,
+		}
+	}
+
+	ImpatientHTTPClient = &http.Client{
+		Timeout:   5 * time.Second,
+		Transport: transport,
+	}
+}

common/config/config.go

@@ -117,10 +117,10 @@ var ValidThemes = map[string]bool{
 // All duration's unit is seconds
 // Shouldn't larger then RateLimitKeyExpirationDuration
 var (
-	GlobalApiRateLimitNum            = env.Int("GLOBAL_API_RATE_LIMIT", 180)
+	GlobalApiRateLimitNum            = env.Int("GLOBAL_API_RATE_LIMIT", 240)
 	GlobalApiRateLimitDuration int64 = 3 * 60
 
-	GlobalWebRateLimitNum            = env.Int("GLOBAL_WEB_RATE_LIMIT", 60)
+	GlobalWebRateLimitNum            = env.Int("GLOBAL_WEB_RATE_LIMIT", 120)
 	GlobalWebRateLimitDuration int64 = 3 * 60
 
 	UploadRateLimitNum            = 10
@@ -144,3 +144,7 @@ var MetricFailChanSize = env.Int("METRIC_FAIL_CHAN_SIZE", 128)
 var InitialRootToken = os.Getenv("INITIAL_ROOT_TOKEN")
 
 var GeminiVersion = env.String("GEMINI_VERSION", "v1")
+
+var RelayProxy = env.String("RELAY_PROXY", "")
+var UserContentRequestProxy = env.String("USER_CONTENT_REQUEST_PROXY", "")
+var UserContentRequestTimeout = env.Int("USER_CONTENT_REQUEST_TIMEOUT", 30)

common/config/key.go

@@ -1,12 +0,0 @@
-package config
-
-const (
-	KeyPrefix = "cfg_"
-
-	KeyAPIVersion = KeyPrefix + "api_version"
-	KeyLibraryID  = KeyPrefix + "library_id"
-	KeyPlugin     = KeyPrefix + "plugin"
-	KeySK         = KeyPrefix + "sk"
-	KeyAK         = KeyPrefix + "ak"
-	KeyRegion     = KeyPrefix + "region"
-)

common/ctxkey/key.go

@@ -1,7 +1,22 @@
 package ctxkey
 
-var (
-	RequestModel     = "request_model"
-	ConvertedRequest = "converted_request"
-	OriginalModel    = "original_model"
+const (
+	Config            = "config"
+	Id                = "id"
+	Username          = "username"
+	Role              = "role"
+	Status            = "status"
+	Channel           = "channel"
+	ChannelId         = "channel_id"
+	SpecificChannelId = "specific_channel_id"
+	RequestModel      = "request_model"
+	ConvertedRequest  = "converted_request"
+	OriginalModel     = "original_model"
+	Group             = "group"
+	ModelMapping      = "model_mapping"
+	ChannelName       = "channel_name"
+	TokenId           = "token_id"
+	TokenName         = "token_name"
+	BaseURL           = "base_url"
+	AvailableModels   = "available_models"
 )

common/helper/helper.go

@@ -2,6 +2,7 @@ package helper
 
 import (
 	"fmt"
+	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/random"
 	"html/template"
 	"log"
@@ -105,6 +106,11 @@ func GenRequestID() string {
 	return GetTimeString() + random.GetRandomNumberString(8)
 }
 
+func GetResponseID(c *gin.Context) string {
+	logID := c.GetString(RequestIdKey)
+	return fmt.Sprintf("chatcmpl-%s", logID)
+}
+
 func Max(a int, b int) int {
 	if a >= b {
 		return a

common/helper/key.go

@@ -0,0 +1,5 @@
+package helper
+
+const (
+	RequestIdKey = "X-Oneapi-Request-Id"
+)

common/image/image.go

@@ -3,6 +3,7 @@ package image
 import (
 	"bytes"
 	"encoding/base64"
+	"github.com/songquanpeng/one-api/common/client"
 	"image"
 	_ "image/gif"
 	_ "image/jpeg"
@@ -16,10 +17,10 @@ import (
 )
 
 // Regex to match data URL pattern
-var	dataURLPattern = regexp.MustCompile(`data:image/([^;]+);base64,(.*)`)
+var dataURLPattern = regexp.MustCompile(`data:image/([^;]+);base64,(.*)`)
 
 func IsImageUrl(url string) (bool, error) {
-	resp, err := http.Head(url)
+	resp, err := client.UserContentRequestHTTPClient.Head(url)
 	if err != nil {
 		return false, err
 	}
@@ -34,7 +35,7 @@ func GetImageSizeFromUrl(url string) (width int, height int, err error) {
 	if !isImage {
 		return
 	}
-	resp, err := http.Get(url)
+	resp, err := client.UserContentRequestHTTPClient.Get(url)
 	if err != nil {
 		return
 	}

common/logger/constants.go

@@ -1,7 +1,3 @@
 package logger
 
-const (
-	RequestIdKey = "X-Oneapi-Request-Id"
-)
-
 var LogDir string

common/logger/logger.go

@@ -3,15 +3,16 @@ package logger
 import (
 	"context"
 	"fmt"
-	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/config"
-	"github.com/songquanpeng/one-api/common/helper"
 	"io"
 	"log"
 	"os"
 	"path/filepath"
 	"sync"
 	"time"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
 )
 
 const (
@@ -21,28 +22,20 @@ const (
 	loggerError = "ERR"
 )
 
-var setupLogLock sync.Mutex
-var setupLogWorking bool
+var setupLogOnce sync.Once
 
 func SetupLogger() {
-	if LogDir != "" {
-		ok := setupLogLock.TryLock()
-		if !ok {
-			log.Println("setup log is already working")
-			return
+	setupLogOnce.Do(func() {
+		if LogDir != "" {
+			logPath := filepath.Join(LogDir, fmt.Sprintf("oneapi-%s.log", time.Now().Format("20060102")))
+			fd, err := os.OpenFile(logPath, os.O_APPEND|os.O_CREATE|os.O_WRONLY, 0644)
+			if err != nil {
+				log.Fatal("failed to open log file")
+			}
+			gin.DefaultWriter = io.MultiWriter(os.Stdout, fd)
+			gin.DefaultErrorWriter = io.MultiWriter(os.Stderr, fd)
 		}
-		defer func() {
-			setupLogLock.Unlock()
-			setupLogWorking = false
-		}()
-		logPath := filepath.Join(LogDir, fmt.Sprintf("oneapi-%s.log", time.Now().Format("20060102")))
-		fd, err := os.OpenFile(logPath, os.O_APPEND|os.O_CREATE|os.O_WRONLY, 0644)
-		if err != nil {
-			log.Fatal("failed to open log file")
-		}
-		gin.DefaultWriter = io.MultiWriter(os.Stdout, fd)
-		gin.DefaultErrorWriter = io.MultiWriter(os.Stderr, fd)
-	}
+	})
 }
 
 func SysLog(s string) {
@@ -50,11 +43,19 @@ func SysLog(s string) {
 	_, _ = fmt.Fprintf(gin.DefaultWriter, "[SYS] %v | %s \n", t.Format("2006/01/02 - 15:04:05"), s)
 }
 
+func SysLogf(format string, a ...any) {
+	SysLog(fmt.Sprintf(format, a...))
+}
+
 func SysError(s string) {
 	t := time.Now()
 	_, _ = fmt.Fprintf(gin.DefaultErrorWriter, "[SYS] %v | %s \n", t.Format("2006/01/02 - 15:04:05"), s)
 }
 
+func SysErrorf(format string, a ...any) {
+	SysError(fmt.Sprintf(format, a...))
+}
+
 func Debug(ctx context.Context, msg string) {
 	if config.DebugEnabled {
 		logHelper(ctx, loggerDEBUG, msg)
@@ -94,18 +95,13 @@ func logHelper(ctx context.Context, level string, msg string) {
 	if level == loggerINFO {
 		writer = gin.DefaultWriter
 	}
-	id := ctx.Value(RequestIdKey)
+	id := ctx.Value(helper.RequestIdKey)
 	if id == nil {
 		id = helper.GenRequestID()
 	}
 	now := time.Now()
 	_, _ = fmt.Fprintf(writer, "[%s] %v | %s | %s \n", level, now.Format("2006/01/02 - 15:04:05"), id, msg)
-	if !setupLogWorking {
-		setupLogWorking = true
-		go func() {
-			SetupLogger()
-		}()
-	}
+	SetupLogger()
 }
 
 func FatalLog(v ...any) {

controller/auth/wechat.go

@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/controller"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
@@ -136,7 +137,7 @@ func WeChatBind(c *gin.Context) {
 		})
 		return
 	}
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	user := model.User{
 		Id: id,
 	}

controller/billing.go

@@ -3,6 +3,7 @@ package controller
 import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/model"
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
 )
@@ -14,13 +15,13 @@ func GetSubscription(c *gin.Context) {
 	var token *model.Token
 	var expiredTime int64
 	if config.DisplayTokenStatEnabled {
-		tokenId := c.GetInt("token_id")
+		tokenId := c.GetInt(ctxkey.TokenId)
 		token, err = model.GetTokenById(tokenId)
 		expiredTime = token.ExpiredTime
 		remainQuota = token.RemainQuota
 		usedQuota = token.UsedQuota
 	} else {
-		userId := c.GetInt("id")
+		userId := c.GetInt(ctxkey.Id)
 		remainQuota, err = model.GetUserQuota(userId)
 		if err != nil {
 			usedQuota, err = model.GetUserUsedQuota(userId)
@@ -64,11 +65,11 @@ func GetUsage(c *gin.Context) {
 	var err error
 	var token *model.Token
 	if config.DisplayTokenStatEnabled {
-		tokenId := c.GetInt("token_id")
+		tokenId := c.GetInt(ctxkey.TokenId)
 		token, err = model.GetTokenById(tokenId)
 		quota = token.UsedQuota
 	} else {
-		userId := c.GetInt("id")
+		userId := c.GetInt(ctxkey.Id)
 		quota, err = model.GetUserUsedQuota(userId)
 	}
 	if err != nil {

controller/channel-billing.go

@@ -4,12 +4,12 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common/client"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/model"
 	"github.com/songquanpeng/one-api/monitor"
 	"github.com/songquanpeng/one-api/relay/channeltype"
-	"github.com/songquanpeng/one-api/relay/client"
 	"io"
 	"net/http"
 	"strconv"

controller/channel-test.go

@@ -5,7 +5,17 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"net/url"
+	"strconv"
+	"strings"
+	"sync"
+	"time"
+
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/common/message"
 	"github.com/songquanpeng/one-api/middleware"
@@ -17,14 +27,6 @@ import (
 	"github.com/songquanpeng/one-api/relay/meta"
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/relaymode"
-	"io"
-	"net/http"
-	"net/http/httptest"
-	"net/url"
-	"strconv"
-	"strings"
-	"sync"
-	"time"
 
 	"github.com/gin-gonic/gin"
 )
@@ -54,8 +56,10 @@ func testChannel(channel *model.Channel) (err error, openaiErr *relaymodel.Error
 	}
 	c.Request.Header.Set("Authorization", "Bearer "+channel.Key)
 	c.Request.Header.Set("Content-Type", "application/json")
-	c.Set("channel", channel.Type)
-	c.Set("base_url", channel.GetBaseURL())
+	c.Set(ctxkey.Channel, channel.Type)
+	c.Set(ctxkey.BaseURL, channel.GetBaseURL())
+	cfg, _ := channel.LoadConfig()
+	c.Set(ctxkey.Config, cfg)
 	middleware.SetupContextForSelectedChannel(c, channel, "")
 	meta := meta.GetByContext(c)
 	apiType := channeltype.ToAPIType(channel.Type)
@@ -64,12 +68,20 @@ func testChannel(channel *model.Channel) (err error, openaiErr *relaymodel.Error
 		return fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), nil
 	}
 	adaptor.Init(meta)
-	modelName := adaptor.GetModelList()[0]
-	if !strings.Contains(channel.Models, modelName) {
+	var modelName string
+	modelList := adaptor.GetModelList()
+	modelMap := channel.GetModelMapping()
+	if len(modelList) != 0 {
+		modelName = modelList[0]
+	}
+	if modelName == "" || !strings.Contains(channel.Models, modelName) {
 		modelNames := strings.Split(channel.Models, ",")
 		if len(modelNames) > 0 {
 			modelName = modelNames[0]
 		}
+		if modelMap != nil && modelMap[modelName] != "" {
+			modelName = modelMap[modelName]
+		}
 	}
 	request := buildTestRequest()
 	request.Model = modelName
@@ -82,6 +94,7 @@ func testChannel(channel *model.Channel) (err error, openaiErr *relaymodel.Error
 	if err != nil {
 		return err, nil
 	}
+	logger.SysLog(string(jsonData))
 	requestBody := bytes.NewBuffer(jsonData)
 	c.Request.Body = io.NopCloser(requestBody)
 	resp, err := adaptor.DoRequest(c, meta, requestBody)

controller/log.go

@@ -3,6 +3,7 @@ package controller
 import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strconv"
@@ -41,7 +42,7 @@ func GetUserLogs(c *gin.Context) {
 	if p < 0 {
 		p = 0
 	}
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	logType, _ := strconv.Atoi(c.Query("type"))
 	startTimestamp, _ := strconv.ParseInt(c.Query("start_timestamp"), 10, 64)
 	endTimestamp, _ := strconv.ParseInt(c.Query("end_timestamp"), 10, 64)
@@ -83,7 +84,7 @@ func SearchAllLogs(c *gin.Context) {
 
 func SearchUserLogs(c *gin.Context) {
 	keyword := c.Query("keyword")
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	logs, err := model.SearchUserLogs(userId, keyword)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -122,7 +123,7 @@ func GetLogsStat(c *gin.Context) {
 }
 
 func GetLogsSelfStat(c *gin.Context) {
-	username := c.GetString("username")
+	username := c.GetString(ctxkey.Username)
 	logType, _ := strconv.Atoi(c.Query("type"))
 	startTimestamp, _ := strconv.ParseInt(c.Query("start_timestamp"), 10, 64)
 	endTimestamp, _ := strconv.ParseInt(c.Query("end_timestamp"), 10, 64)

controller/model.go

@@ -3,6 +3,7 @@ package controller
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/model"
 	relay "github.com/songquanpeng/one-api/relay"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
@@ -131,10 +132,10 @@ func ListAllModels(c *gin.Context) {
 func ListModels(c *gin.Context) {
 	ctx := c.Request.Context()
 	var availableModels []string
-	if c.GetString("available_models") != "" {
-		availableModels = strings.Split(c.GetString("available_models"), ",")
+	if c.GetString(ctxkey.AvailableModels) != "" {
+		availableModels = strings.Split(c.GetString(ctxkey.AvailableModels), ",")
 	} else {
-		userId := c.GetInt("id")
+		userId := c.GetInt(ctxkey.Id)
 		userGroup, _ := model.CacheGetUserGroup(userId)
 		availableModels, _ = model.CacheGetGroupModels(ctx, userGroup)
 	}
@@ -186,7 +187,7 @@ func RetrieveModel(c *gin.Context) {
 
 func GetUserAvailableModels(c *gin.Context) {
 	ctx := c.Request.Context()
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	userGroup, err := model.CacheGetUserGroup(id)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{

controller/redemption.go

@@ -3,6 +3,7 @@ package controller
 import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/random"
 	"github.com/songquanpeng/one-api/model"
@@ -109,7 +110,7 @@ func AddRedemption(c *gin.Context) {
 	for i := 0; i < redemption.Count; i++ {
 		key := random.GetUUID()
 		cleanRedemption := model.Redemption{
-			UserId:      c.GetInt("id"),
+			UserId:      c.GetInt(ctxkey.Id),
 			Name:        redemption.Name,
 			Key:         key,
 			CreatedTime: helper.GetTimestamp(),

controller/relay.go

@@ -4,6 +4,9 @@ import (
 	"bytes"
 	"context"
 	"fmt"
+	"io"
+	"net/http"
+
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
@@ -16,8 +19,6 @@ import (
 	"github.com/songquanpeng/one-api/relay/controller"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/relaymode"
-	"io"
-	"net/http"
 )
 
 // https://platform.openai.com/docs/api-reference/chat
@@ -46,18 +47,19 @@ func Relay(c *gin.Context) {
 		requestBody, _ := common.GetRequestBody(c)
 		logger.Debugf(ctx, "request body: %s", string(requestBody))
 	}
-	channelId := c.GetInt("channel_id")
+	channelId := c.GetInt(ctxkey.ChannelId)
+	userId := c.GetInt("id")
 	bizErr := relayHelper(c, relayMode)
 	if bizErr == nil {
 		monitor.Emit(channelId, true)
 		return
 	}
 	lastFailedChannelId := channelId
-	channelName := c.GetString("channel_name")
-	group := c.GetString("group")
+	channelName := c.GetString(ctxkey.ChannelName)
+	group := c.GetString(ctxkey.Group)
 	originalModel := c.GetString(ctxkey.OriginalModel)
-	go processChannelRelayError(ctx, channelId, channelName, bizErr)
-	requestId := c.GetString(logger.RequestIdKey)
+	go processChannelRelayError(ctx, userId, channelId, channelName, bizErr)
+	requestId := c.GetString(helper.RequestIdKey)
 	retryTimes := config.RetryTimes
 	if !shouldRetry(c, bizErr.StatusCode) {
 		logger.Errorf(ctx, "relay error happen, status code is %d, won't retry in this case", bizErr.StatusCode)
@@ -80,10 +82,10 @@ func Relay(c *gin.Context) {
 		if bizErr == nil {
 			return
 		}
-		channelId := c.GetInt("channel_id")
+		channelId := c.GetInt(ctxkey.ChannelId)
 		lastFailedChannelId = channelId
-		channelName := c.GetString("channel_name")
-		go processChannelRelayError(ctx, channelId, channelName, bizErr)
+		channelName := c.GetString(ctxkey.ChannelName)
+		go processChannelRelayError(ctx, userId, channelId, channelName, bizErr)
 	}
 	if bizErr != nil {
 		if bizErr.StatusCode == http.StatusTooManyRequests {
@@ -97,7 +99,7 @@ func Relay(c *gin.Context) {
 }
 
 func shouldRetry(c *gin.Context, statusCode int) bool {
-	if _, ok := c.Get("specific_channel_id"); ok {
+	if _, ok := c.Get(ctxkey.SpecificChannelId); ok {
 		return false
 	}
 	if statusCode == http.StatusTooManyRequests {
@@ -115,8 +117,8 @@ func shouldRetry(c *gin.Context, statusCode int) bool {
 	return true
 }
 
-func processChannelRelayError(ctx context.Context, channelId int, channelName string, err *model.ErrorWithStatusCode) {
-	logger.Errorf(ctx, "relay error (channel #%d): %s", channelId, err.Message)
+func processChannelRelayError(ctx context.Context, userId int, channelId int, channelName string, err *model.ErrorWithStatusCode) {
+	logger.Errorf(ctx, "relay error (channel id %d, user id: %d): %s", channelId, userId, err.Message)
 	// https://platform.openai.com/docs/guides/error-codes/api-errors
 	if monitor.ShouldDisableChannel(&err.Error, err.StatusCode) {
 		monitor.DisableChannel(channelId, channelName, err.Message)

controller/token.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/network"
 	"github.com/songquanpeng/one-api/common/random"
@@ -13,7 +14,7 @@ import (
 )
 
 func GetAllTokens(c *gin.Context) {
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	p, _ := strconv.Atoi(c.Query("p"))
 	if p < 0 {
 		p = 0
@@ -38,7 +39,7 @@ func GetAllTokens(c *gin.Context) {
 }
 
 func SearchTokens(c *gin.Context) {
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	keyword := c.Query("keyword")
 	tokens, err := model.SearchUserTokens(userId, keyword)
 	if err != nil {
@@ -58,7 +59,7 @@ func SearchTokens(c *gin.Context) {
 
 func GetToken(c *gin.Context) {
 	id, err := strconv.Atoi(c.Param("id"))
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -83,8 +84,8 @@ func GetToken(c *gin.Context) {
 }
 
 func GetTokenStatus(c *gin.Context) {
-	tokenId := c.GetInt("token_id")
-	userId := c.GetInt("id")
+	tokenId := c.GetInt(ctxkey.TokenId)
+	userId := c.GetInt(ctxkey.Id)
 	token, err := model.GetTokenByIds(tokenId, userId)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -139,7 +140,7 @@ func AddToken(c *gin.Context) {
 	}
 
 	cleanToken := model.Token{
-		UserId:         c.GetInt("id"),
+		UserId:         c.GetInt(ctxkey.Id),
 		Name:           token.Name,
 		Key:            random.GenerateKey(),
 		CreatedTime:    helper.GetTimestamp(),
@@ -168,7 +169,7 @@ func AddToken(c *gin.Context) {
 
 func DeleteToken(c *gin.Context) {
 	id, _ := strconv.Atoi(c.Param("id"))
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	err := model.DeleteTokenById(id, userId)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -185,7 +186,7 @@ func DeleteToken(c *gin.Context) {
 }
 
 func UpdateToken(c *gin.Context) {
-	userId := c.GetInt("id")
+	userId := c.GetInt(ctxkey.Id)
 	statusOnly := c.Query("status_only")
 	token := model.Token{}
 	err := c.ShouldBindJSON(&token)

controller/user.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/random"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
@@ -172,6 +173,7 @@ func Register(c *gin.Context) {
 		})
 		return
 	}
+
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
@@ -238,7 +240,7 @@ func GetUser(c *gin.Context) {
 		})
 		return
 	}
-	myRole := c.GetInt("role")
+	myRole := c.GetInt(ctxkey.Role)
 	if myRole <= user.Role && myRole != model.RoleRootUser {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -255,7 +257,7 @@ func GetUser(c *gin.Context) {
 }
 
 func GetUserDashboard(c *gin.Context) {
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	now := time.Now()
 	startOfDay := now.Truncate(24*time.Hour).AddDate(0, 0, -6).Unix()
 	endOfDay := now.Truncate(24 * time.Hour).Add(24*time.Hour - time.Second).Unix()
@@ -278,7 +280,7 @@ func GetUserDashboard(c *gin.Context) {
 }
 
 func GenerateAccessToken(c *gin.Context) {
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	user, err := model.GetUserById(id, true)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -314,7 +316,7 @@ func GenerateAccessToken(c *gin.Context) {
 }
 
 func GetAffCode(c *gin.Context) {
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	user, err := model.GetUserById(id, true)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -342,7 +344,7 @@ func GetAffCode(c *gin.Context) {
 }
 
 func GetSelf(c *gin.Context) {
-	id := c.GetInt("id")
+	id := c.GetInt(ctxkey.Id)
 	user, err := model.GetUserById(id, false)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
@@ -387,7 +389,7 @@ func UpdateUser(c *gin.Context) {
 		})
 		return
 	}
-	myRole := c.GetInt("role")
+	myRole := c.GetInt(ctxkey.Role)
 	if myRole <= originUser.Role && myRole != model.RoleRootUser {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -445,7 +447,7 @@ func UpdateSelf(c *gin.Context) {
 	}
 
 	cleanUser := model.User{
-		Id:          c.GetInt("id"),
+		Id:          c.GetInt(ctxkey.Id),
 		Username:    user.Username,
 		Password:    user.Password,
 		DisplayName: user.DisplayName,

go.mod

@@ -4,42 +4,42 @@ module github.com/songquanpeng/one-api
 go 1.20
 
 require (
-	github.com/aws/aws-sdk-go-v2 v1.26.1
-	github.com/aws/aws-sdk-go-v2/credentials v1.17.11
-	github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4
-	github.com/gin-contrib/cors v1.7.1
-	github.com/gin-contrib/gzip v1.0.0
-	github.com/gin-contrib/sessions v1.0.0
-	github.com/gin-contrib/static v1.1.1
-	github.com/gin-gonic/gin v1.9.1
-	github.com/go-playground/validator/v10 v10.19.0
+	github.com/aws/aws-sdk-go-v2 v1.27.0
+	github.com/aws/aws-sdk-go-v2/credentials v1.17.15
+	github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.8.3
+	github.com/gin-contrib/cors v1.7.2
+	github.com/gin-contrib/gzip v1.0.1
+	github.com/gin-contrib/sessions v1.0.1
+	github.com/gin-contrib/static v1.1.2
+	github.com/gin-gonic/gin v1.10.0
+	github.com/go-playground/validator/v10 v10.20.0
 	github.com/go-redis/redis/v8 v8.11.5
 	github.com/golang-jwt/jwt v3.2.2+incompatible
 	github.com/google/uuid v1.6.0
 	github.com/gorilla/websocket v1.5.1
 	github.com/jinzhu/copier v0.4.0
 	github.com/pkg/errors v0.9.1
-	github.com/pkoukk/tiktoken-go v0.1.6
+	github.com/pkoukk/tiktoken-go v0.1.7
 	github.com/smartystreets/goconvey v1.8.1
 	github.com/stretchr/testify v1.9.0
-	golang.org/x/crypto v0.22.0
-	golang.org/x/image v0.15.0
+	golang.org/x/crypto v0.23.0
+	golang.org/x/image v0.16.0
 	gorm.io/driver/mysql v1.5.6
 	gorm.io/driver/postgres v1.5.7
 	gorm.io/driver/sqlite v1.5.5
-	gorm.io/gorm v1.25.9
+	gorm.io/gorm v1.25.10
 )
 
 require (
 	filippo.io/edwards25519 v1.1.0 // indirect
 	github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2 // indirect
-	github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5 // indirect
-	github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.7 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.7 // indirect
 	github.com/aws/smithy-go v1.20.2 // indirect
-	github.com/bytedance/sonic v1.11.5 // indirect
+	github.com/bytedance/sonic v1.11.6 // indirect
 	github.com/bytedance/sonic/loader v0.1.1 // indirect
 	github.com/cespare/xxhash/v2 v2.3.0 // indirect
-	github.com/cloudwego/base64x v0.1.3 // indirect
+	github.com/cloudwego/base64x v0.1.4 // indirect
 	github.com/cloudwego/iasm v0.2.0 // indirect
 	github.com/davecgh/go-spew v1.1.1 // indirect
 	github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f // indirect
@@ -50,7 +50,7 @@ require (
 	github.com/go-playground/locales v0.14.1 // indirect
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-sql-driver/mysql v1.8.1 // indirect
-	github.com/goccy/go-json v0.10.2 // indirect
+	github.com/goccy/go-json v0.10.3 // indirect
 	github.com/gopherjs/gopherjs v1.17.2 // indirect
 	github.com/gorilla/context v1.1.2 // indirect
 	github.com/gorilla/securecookie v1.1.2 // indirect
@@ -70,16 +70,16 @@ require (
 	github.com/mattn/go-sqlite3 v2.0.3+incompatible // indirect
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
 	github.com/modern-go/reflect2 v1.0.2 // indirect
-	github.com/pelletier/go-toml/v2 v2.2.1 // indirect
+	github.com/pelletier/go-toml/v2 v2.2.2 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	github.com/smarty/assertions v1.15.0 // indirect
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.12 // indirect
-	golang.org/x/arch v0.7.0 // indirect
-	golang.org/x/net v0.24.0 // indirect
+	golang.org/x/arch v0.8.0 // indirect
+	golang.org/x/net v0.25.0 // indirect
 	golang.org/x/sync v0.7.0 // indirect
-	golang.org/x/sys v0.19.0 // indirect
-	golang.org/x/text v0.14.0 // indirect
-	google.golang.org/protobuf v1.33.0 // indirect
+	golang.org/x/sys v0.20.0 // indirect
+	golang.org/x/text v0.15.0 // indirect
+	google.golang.org/protobuf v1.34.1 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )

go.sum

@@ -2,20 +2,32 @@ filippo.io/edwards25519 v1.1.0 h1:FNf4tywRC1HmFuKW5xopWpigGjJKiJSV0Cqo0cJWDaA=
 filippo.io/edwards25519 v1.1.0/go.mod h1:BxyFTGdWcka3PhytdK4V28tE5sGfRvvvRV7EaN4VDT4=
 github.com/aws/aws-sdk-go-v2 v1.26.1 h1:5554eUqIYVWpU0YmeeYZ0wU64H2VLBs8TlhRB2L+EkA=
 github.com/aws/aws-sdk-go-v2 v1.26.1/go.mod h1:ffIFB97e2yNsv4aTSGkqtHnppsIJzw7G7BReUZ3jCXM=
+github.com/aws/aws-sdk-go-v2 v1.27.0 h1:7bZWKoXhzI+mMR/HjdMx8ZCC5+6fY0lS5tr0bbgiLlo=
+github.com/aws/aws-sdk-go-v2 v1.27.0/go.mod h1:ffIFB97e2yNsv4aTSGkqtHnppsIJzw7G7BReUZ3jCXM=
 github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2 h1:x6xsQXGSmW6frevwDA+vi/wqhp1ct18mVXYN08/93to=
 github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2/go.mod h1:lPprDr1e6cJdyYeGXnRaJoP4Md+cDBvi2eOj00BlGmg=
 github.com/aws/aws-sdk-go-v2/credentials v1.17.11 h1:YuIB1dJNf1Re822rriUOTxopaHHvIq0l/pX3fwO+Tzs=
 github.com/aws/aws-sdk-go-v2/credentials v1.17.11/go.mod h1:AQtFPsDH9bI2O+71anW6EKL+NcD7LG3dpKGMV4SShgo=
+github.com/aws/aws-sdk-go-v2/credentials v1.17.15 h1:YDexlvDRCA8ems2T5IP1xkMtOZ1uLJOCJdTr0igs5zo=
+github.com/aws/aws-sdk-go-v2/credentials v1.17.15/go.mod h1:vxHggqW6hFNaeNC0WyXS3VdyjcV0a4KMUY4dKJ96buU=
 github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5 h1:aw39xVGeRWlWx9EzGVnhOR4yOjQDHPQ6o6NmBlscyQg=
 github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5/go.mod h1:FSaRudD0dXiMPK2UjknVwwTYyZMRsHv3TtkabsZih5I=
+github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.7 h1:lf/8VTF2cM+N4SLzaYJERKEWAXq8MOMpZfU6wEPWsPk=
+github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.7/go.mod h1:4SjkU7QiqK2M9oozyMzfZ/23LmUY+h3oFqhdeP5OMiI=
 github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5 h1:PG1F3OD1szkuQPzDw3CIQsRIrtTlUC3lP84taWzHlq0=
 github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5/go.mod h1:jU1li6RFryMz+so64PpKtudI+QzbKoIEivqdf6LNpOc=
+github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.7 h1:4OYVp0705xu8yjdyoWix0r9wPIRXnIzzOoUpQVHIJ/g=
+github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.7/go.mod h1:vd7ESTEvI76T2Na050gODNmNU7+OyKrIKroYTu4ABiI=
 github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4 h1:JgHnonzbnA3pbqj76wYsSZIZZQYBxkmMEjvL6GHy8XU=
 github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4/go.mod h1:nZspkhg+9p8iApLFoyAqfyuMP0F38acy2Hm3r5r95Cg=
+github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.8.3 h1:Fihjyd6DeNjcawBEGLH9dkIEUi6AdhucDKPE9nJ4QiY=
+github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.8.3/go.mod h1:opvUj3ismqSCxYc+m4WIjPL0ewZGtvp0ess7cKvBPOQ=
 github.com/aws/smithy-go v1.20.2 h1:tbp628ireGtzcHDDmLT/6ADHidqnwgF57XOXZe6tp4Q=
 github.com/aws/smithy-go v1.20.2/go.mod h1:krry+ya/rV9RDcV/Q16kpu6ypI4K2czasz0NC3qS14E=
 github.com/bytedance/sonic v1.11.5 h1:G00FYjjqll5iQ1PYXynbg/hyzqBqavH8Mo9/oTopd9k=
 github.com/bytedance/sonic v1.11.5/go.mod h1:X2PC2giUdj/Cv2lliWFLk6c/DUQok5rViJSemeB0wDw=
+github.com/bytedance/sonic v1.11.6 h1:oUp34TzMlL+OY1OUWxHqsdkgC/Zfc85zGqw9siXjrc0=
+github.com/bytedance/sonic v1.11.6/go.mod h1:LysEHSvpvDySVdC2f87zGWf6CIKJcAvqab1ZaiQtds4=
 github.com/bytedance/sonic/loader v0.1.0/go.mod h1:UmRT+IRTGKz/DAkzcEGzyVqQFJ7H9BqwBO3pm9H/+HY=
 github.com/bytedance/sonic/loader v0.1.1 h1:c+e5Pt1k/cy5wMveRDyk2X4B9hF4g7an8N3zCYjJFNM=
 github.com/bytedance/sonic/loader v0.1.1/go.mod h1:ncP89zfokxS5LZrJxl5z0UJcsk4M4yY2JpfqGeCtNLU=
@@ -23,6 +35,8 @@ github.com/cespare/xxhash/v2 v2.3.0 h1:UL815xU9SqsFlibzuggzjXhog7bL6oX9BbNZnL2UF
 github.com/cespare/xxhash/v2 v2.3.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
 github.com/cloudwego/base64x v0.1.3 h1:b5J/l8xolB7dyDTTmhJP2oTs5LdrjyrUFuNxdfq5hAg=
 github.com/cloudwego/base64x v0.1.3/go.mod h1:1+1K5BUHIQzyapgpF7LwvOGAEDicKtt1umPV+aN8pi8=
+github.com/cloudwego/base64x v0.1.4 h1:jwCgWpFanWmN8xoIUHa2rtzmkd5J2plF/dnLS6Xd/0Y=
+github.com/cloudwego/base64x v0.1.4/go.mod h1:0zlkT4Wn5C6NdauXdJRhSKRlJvmclQ1hhJgA0rcu/8w=
 github.com/cloudwego/iasm v0.2.0 h1:1KNIy1I1H9hNNFEEH3DVnI4UujN+1zjpuk6gwHLTssg=
 github.com/cloudwego/iasm v0.2.0/go.mod h1:8rXZaNYT2n95jn+zTI1sDr+IgcD2GVs0nlbbQPiEFhY=
 github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
@@ -39,16 +53,26 @@ github.com/gabriel-vasile/mimetype v1.4.3 h1:in2uUcidCuFcDKtdcBxlR0rJ1+fsokWf+uq
 github.com/gabriel-vasile/mimetype v1.4.3/go.mod h1:d8uq/6HKRL6CGdk+aubisF/M5GcPfT7nKyLpA0lbSSk=
 github.com/gin-contrib/cors v1.7.1 h1:s9SIppU/rk8enVvkzwiC2VK3UZ/0NNGsWfUKvV55rqs=
 github.com/gin-contrib/cors v1.7.1/go.mod h1:n/Zj7B4xyrgk/cX1WCX2dkzFfaNm/xJb6oIUk7WTtps=
+github.com/gin-contrib/cors v1.7.2 h1:oLDHxdg8W/XDoN/8zamqk/Drgt4oVZDvaV0YmvVICQw=
+github.com/gin-contrib/cors v1.7.2/go.mod h1:SUJVARKgQ40dmrzgXEVxj2m7Ig1v1qIboQkPDTQ9t2E=
 github.com/gin-contrib/gzip v1.0.0 h1:UKN586Po/92IDX6ie5CWLgMI81obiIp5nSP85T3wlTk=
 github.com/gin-contrib/gzip v1.0.0/go.mod h1:CtG7tQrPB3vIBo6Gat9FVUsis+1emjvQqd66ME5TdnE=
+github.com/gin-contrib/gzip v1.0.1 h1:HQ8ENHODeLY7a4g1Au/46Z92bdGFl74OhxcZble9WJE=
+github.com/gin-contrib/gzip v1.0.1/go.mod h1:njt428fdUNRvjuJf16tZMYZ2Yl+WQB53X5wmhDwXvC4=
 github.com/gin-contrib/sessions v1.0.0 h1:r5GLta4Oy5xo9rAwMHx8B4wLpeRGHMdz9NafzJAdP8Y=
 github.com/gin-contrib/sessions v1.0.0/go.mod h1:DN0f4bvpqMQElDdi+gNGScrP2QEI04IErRyMFyorUOI=
+github.com/gin-contrib/sessions v1.0.1 h1:3hsJyNs7v7N8OtelFmYXFrulAf6zSR7nW/putcPEHxI=
+github.com/gin-contrib/sessions v1.0.1/go.mod h1:ouxSFM24/OgIud5MJYQJLpy6AwxQ5EYO9yLhbtObGkM=
 github.com/gin-contrib/sse v0.1.0 h1:Y/yl/+YNO8GZSjAhjMsSuLt29uWRFHdHYUb5lYOV9qE=
 github.com/gin-contrib/sse v0.1.0/go.mod h1:RHrZQHXnP2xjPF+u1gW/2HnVO7nvIa9PG3Gm+fLHvGI=
 github.com/gin-contrib/static v1.1.1 h1:XEvBd4DDLG1HBlyPBQU1XO8NlTpw6mgdqcPteetYA5k=
 github.com/gin-contrib/static v1.1.1/go.mod h1:yRGmar7+JYvbMLRPIi4H5TVVSBwULfT9vetnVD0IO74=
+github.com/gin-contrib/static v1.1.2 h1:c3kT4bFkUJn2aoRU3s6XnMjJT8J6nNWJkR0NglqmlZ4=
+github.com/gin-contrib/static v1.1.2/go.mod h1:Fw90ozjHCmZBWbgrsqrDvO28YbhKEKzKp8GixhR4yLw=
 github.com/gin-gonic/gin v1.9.1 h1:4idEAncQnU5cB7BeOkPtxjfCSye0AAm1R0RVIqJ+Jmg=
 github.com/gin-gonic/gin v1.9.1/go.mod h1:hPrL7YrpYKXt5YId3A/Tnip5kqbEAP+KLuI3SUcPTeU=
+github.com/gin-gonic/gin v1.10.0 h1:nTuyha1TYqgedzytsKYqna+DfLos46nTv2ygFy86HFU=
+github.com/gin-gonic/gin v1.10.0/go.mod h1:4PMNQiOhvDRa013RKVbsiNwoyezlm2rm0uX/T7kzp5Y=
 github.com/go-playground/assert/v2 v2.2.0 h1:JvknZsQTYeFEAhQwI4qEt9cyV5ONwRHC+lYKSsYSR8s=
 github.com/go-playground/locales v0.14.1 h1:EWaQ/wswjilfKLTECiXz7Rh+3BjFhfDFKv/oXslEjJA=
 github.com/go-playground/locales v0.14.1/go.mod h1:hxrqLVvrK65+Rwrd5Fc6F2O76J/NuW9t0sjnWqG1slY=
@@ -56,6 +80,8 @@ github.com/go-playground/universal-translator v0.18.1 h1:Bcnm0ZwsGyWbCzImXv+pAJn
 github.com/go-playground/universal-translator v0.18.1/go.mod h1:xekY+UJKNuX9WP91TpwSH2VMlDf28Uj24BCp08ZFTUY=
 github.com/go-playground/validator/v10 v10.19.0 h1:ol+5Fu+cSq9JD7SoSqe04GMI92cbn0+wvQ3bZ8b/AU4=
 github.com/go-playground/validator/v10 v10.19.0/go.mod h1:dbuPbCMFw/DrkbEynArYaCwl3amGuJotoKCe95atGMM=
+github.com/go-playground/validator/v10 v10.20.0 h1:K9ISHbSaI0lyB2eWMPJo+kOS/FBExVwjEviJTixqxL8=
+github.com/go-playground/validator/v10 v10.20.0/go.mod h1:dbuPbCMFw/DrkbEynArYaCwl3amGuJotoKCe95atGMM=
 github.com/go-redis/redis/v8 v8.11.5 h1:AcZZR7igkdvfVmQTPnu9WE37LRrO/YrBH5zWyjDC0oI=
 github.com/go-redis/redis/v8 v8.11.5/go.mod h1:gREzHqY1hg6oD9ngVRbLStwAWKhA0FEgq8Jd4h5lpwo=
 github.com/go-sql-driver/mysql v1.7.0/go.mod h1:OXbVy3sEdcQ2Doequ6Z5BW6fXNQTmx+9S1MCJN5yJMI=
@@ -63,6 +89,8 @@ github.com/go-sql-driver/mysql v1.8.1 h1:LedoTUt/eveggdHS9qUFC1EFSa8bU2+1pZjSRpv
 github.com/go-sql-driver/mysql v1.8.1/go.mod h1:wEBSXgmK//2ZFJyE+qWnIsVGmvmEKlqwuVSjsCm7DZg=
 github.com/goccy/go-json v0.10.2 h1:CrxCmQqYDkv1z7lO7Wbh2HN93uovUHgrECaO5ZrCXAU=
 github.com/goccy/go-json v0.10.2/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
+github.com/goccy/go-json v0.10.3 h1:KZ5WoDbxAIgm2HNbYckL0se1fHD6rz5j4ywS6ebzDqA=
+github.com/goccy/go-json v0.10.3/go.mod h1:oq7eo15ShAhp70Anwd5lgX2pLfOS3QCiwU/PULtXL6M=
 github.com/golang-jwt/jwt v3.2.2+incompatible h1:IfV12K8xAKAnZqdXVzCZ+TOjboZ2keLg81eXfW3O+oY=
 github.com/golang-jwt/jwt v3.2.2+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
 github.com/google/go-cmp v0.5.6 h1:BKbKCqvP6I+rmFHt06ZmyQtvB8xAkWdhFyr0ZUNZcxQ=
@@ -121,10 +149,14 @@ github.com/onsi/ginkgo v1.16.5 h1:8xi0RTUf59SOSfEtZMvwTvXYMzG4gV23XVHOZiXNtnE=
 github.com/onsi/gomega v1.18.1 h1:M1GfJqGRrBrrGGsbxzV5dqM2U2ApXefZCQpkukxYRLE=
 github.com/pelletier/go-toml/v2 v2.2.1 h1:9TA9+T8+8CUCO2+WYnDLCgrYi9+omqKXyjDtosvtEhg=
 github.com/pelletier/go-toml/v2 v2.2.1/go.mod h1:1t835xjRzz80PqgE6HHgN2JOsmgYu/h4qDAS4n929Rs=
+github.com/pelletier/go-toml/v2 v2.2.2 h1:aYUidT7k73Pcl9nb2gScu7NSrKCSHIDE89b3+6Wq+LM=
+github.com/pelletier/go-toml/v2 v2.2.2/go.mod h1:1t835xjRzz80PqgE6HHgN2JOsmgYu/h4qDAS4n929Rs=
 github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
 github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkoukk/tiktoken-go v0.1.6 h1:JF0TlJzhTbrI30wCvFuiw6FzP2+/bR+FIxUdgEAcUsw=
 github.com/pkoukk/tiktoken-go v0.1.6/go.mod h1:9NiV+i9mJKGj1rYOT+njbv+ZwA/zJxYdewGl6qVatpg=
+github.com/pkoukk/tiktoken-go v0.1.7 h1:qOBHXX4PHtvIvmOtyg1EeKlwFRiMKAcoMp4Q+bLQDmw=
+github.com/pkoukk/tiktoken-go v0.1.7/go.mod h1:9NiV+i9mJKGj1rYOT+njbv+ZwA/zJxYdewGl6qVatpg=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/rogpeppe/go-internal v1.8.0 h1:FCbCCtXNOY3UtUuHUYaghJg4y7Fd14rXifAYUAtL9R8=
@@ -151,23 +183,37 @@ github.com/ugorji/go/codec v1.2.12/go.mod h1:UNopzCgEMSXjBc6AOMqYvWC1ktqTAfzJZUZ
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/arch v0.7.0 h1:pskyeJh/3AmoQ8CPE95vxHLqp1G1GfGNXTmcl9NEKTc=
 golang.org/x/arch v0.7.0/go.mod h1:FEVrYAQjsQXMVJ1nsMoVVXPZg6p2JE2mx8psSWTDQys=
+golang.org/x/arch v0.8.0 h1:3wRIsP3pM4yUptoR96otTUOXI367OS0+c9eeRi9doIc=
+golang.org/x/arch v0.8.0/go.mod h1:FEVrYAQjsQXMVJ1nsMoVVXPZg6p2JE2mx8psSWTDQys=
 golang.org/x/crypto v0.22.0 h1:g1v0xeRhjcugydODzvb3mEM9SQ0HGp9s/nh3COQ/C30=
 golang.org/x/crypto v0.22.0/go.mod h1:vr6Su+7cTlO45qkww3VDJlzDn0ctJvRgYbC2NvXHt+M=
+golang.org/x/crypto v0.23.0 h1:dIJU/v2J8Mdglj/8rJ6UUOM3Zc9zLZxVZwwxMooUSAI=
+golang.org/x/crypto v0.23.0/go.mod h1:CKFgDieR+mRhux2Lsu27y0fO304Db0wZe70UKqHu0v8=
 golang.org/x/image v0.15.0 h1:kOELfmgrmJlw4Cdb7g/QGuB3CvDrXbqEIww/pNtNBm8=
 golang.org/x/image v0.15.0/go.mod h1:HUYqC05R2ZcZ3ejNQsIHQDQiwWM4JBqmm6MKANTp4LE=
+golang.org/x/image v0.16.0 h1:9kloLAKhUufZhA12l5fwnx2NZW39/we1UhBesW433jw=
+golang.org/x/image v0.16.0/go.mod h1:ugSZItdV4nOxyqp56HmXwH0Ry0nBCpjnZdpDaIHdoPs=
 golang.org/x/net v0.24.0 h1:1PcaxkF854Fu3+lvBIx5SYn9wRlBzzcnHZSiaFFAb0w=
 golang.org/x/net v0.24.0/go.mod h1:2Q7sJY5mzlzWjKtYUEXSlBWCdyaioyXzRB2RtU8KVE8=
+golang.org/x/net v0.25.0 h1:d/OCCoBEUq33pjydKrGQhw7IlUPI2Oylr+8qLx49kac=
+golang.org/x/net v0.25.0/go.mod h1:JkAGAh7GEvH74S6FOH42FLoXpXbE/aqXSrIQjXgsiwM=
 golang.org/x/sync v0.7.0 h1:YsImfSBoP9QPYL0xyKJPq0gcaJdG3rInoqxTWbfQu9M=
 golang.org/x/sync v0.7.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
 golang.org/x/sys v0.5.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.19.0 h1:q5f1RH2jigJ1MoAWp2KTp3gm5zAGFUTarQZ5U386+4o=
 golang.org/x/sys v0.19.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
+golang.org/x/sys v0.20.0 h1:Od9JTbYCk261bKm4M/mw7AklTlFYIa0bIp9BgSm1S8Y=
+golang.org/x/sys v0.20.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
 golang.org/x/text v0.14.0 h1:ScX5w1eTa3QqT8oi6+ziP7dTV1S2+ALU0bI+0zXKWiQ=
 golang.org/x/text v0.14.0/go.mod h1:18ZOQIKpY8NJVqYksKHtTdi31H5itFRjB5/qKTNYzSU=
+golang.org/x/text v0.15.0 h1:h1V/4gjBv8v9cjcR6+AR5+/cIYK5N/WAgiv4xlsEtAk=
+golang.org/x/text v0.15.0/go.mod h1:18ZOQIKpY8NJVqYksKHtTdi31H5itFRjB5/qKTNYzSU=
 golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 h1:go1bK/D/BFZV2I8cIQd1NKEZ+0owSTG1fDTci4IqFcE=
 google.golang.org/protobuf v1.33.0 h1:uNO2rsAINq/JlFpSdYEKIZ0uKD/R9cpdv0T+yoGwGmI=
 google.golang.org/protobuf v1.33.0/go.mod h1:c6P6GXX6sHbq/GpV6MGZEdwhWPcYBgnhAHhKbcUYpos=
+google.golang.org/protobuf v1.34.1 h1:9ddQBjfCyZPOHPUiPxpYESBLc+T8P3E+Vo4IbKZgFWg=
+google.golang.org/protobuf v1.34.1/go.mod h1:c6P6GXX6sHbq/GpV6MGZEdwhWPcYBgnhAHhKbcUYpos=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
 gopkg.in/tomb.v1 v1.0.0-20141024135613-dd632973f1e7 h1:uRGJdciOHaEIrze2W8Q3AKkepLTh2hOroT7a+7czfdQ=
@@ -184,5 +230,7 @@ gorm.io/driver/sqlite v1.5.5/go.mod h1:6NgQ7sQWAIFsPrJJl1lSNSu2TABh0ZZ/zm5fosATa
 gorm.io/gorm v1.25.7/go.mod h1:hbnx/Oo0ChWMn1BIhpy1oYozzpM15i4YPuHDmfYtwg8=
 gorm.io/gorm v1.25.9 h1:wct0gxZIELDk8+ZqF/MVnHLkA1rvYlBWUMv2EdsK1g8=
 gorm.io/gorm v1.25.9/go.mod h1:hbnx/Oo0ChWMn1BIhpy1oYozzpM15i4YPuHDmfYtwg8=
+gorm.io/gorm v1.25.10 h1:dQpO+33KalOA+aFYGlK+EfxcI5MbO7EP2yYygwh9h+s=
+gorm.io/gorm v1.25.10/go.mod h1:hbnx/Oo0ChWMn1BIhpy1oYozzpM15i4YPuHDmfYtwg8=
 nullprogram.com/x/optparse v1.0.0/go.mod h1:KdyPE+Igbe0jQUrVfMqDMeJQIJZEuyV7pjYmp6pbG50=
 rsc.io/pdf v0.1.1/go.mod h1:n8OzWcQ6Sp37PL01nO98y4iUCRdTGarVfzxY20ICaU4=

main.go

@@ -7,6 +7,7 @@ import (
 	"github.com/gin-contrib/sessions/cookie"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/client"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/controller"
@@ -23,7 +24,7 @@ var buildFS embed.FS
 
 func main() {
 	logger.SetupLogger()
-	logger.SysLog(fmt.Sprintf("One API %s started", common.Version))
+	logger.SysLogf("One API %s started", common.Version)
 	if os.Getenv("GIN_MODE") != "debug" {
 		gin.SetMode(gin.ReleaseMode)
 	}
@@ -71,7 +72,7 @@ func main() {
 	}
 	if config.MemoryCacheEnabled {
 		logger.SysLog("memory cache enabled")
-		logger.SysError(fmt.Sprintf("sync frequency: %d seconds", config.SyncFrequency))
+		logger.SysLog(fmt.Sprintf("sync frequency: %d seconds", config.SyncFrequency))
 		model.InitChannelCache()
 	}
 	if config.MemoryCacheEnabled {
@@ -94,6 +95,7 @@ func main() {
 		logger.SysLog("metric enabled, will disable channel if too much request failed")
 	}
 	openai.InitTokenEncoders()
+	client.Init()
 
 	// Initialize HTTP server
 	server := gin.New()

middleware/auth.go

@@ -5,6 +5,7 @@ import (
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/blacklist"
+	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/network"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
@@ -120,20 +121,20 @@ func TokenAuth() func(c *gin.Context) {
 			abortWithMessage(c, http.StatusBadRequest, err.Error())
 			return
 		}
-		c.Set("request_model", requestModel)
+		c.Set(ctxkey.RequestModel, requestModel)
 		if token.Models != nil && *token.Models != "" {
-			c.Set("available_models", *token.Models)
+			c.Set(ctxkey.AvailableModels, *token.Models)
 			if requestModel != "" && !isModelInList(requestModel, *token.Models) {
 				abortWithMessage(c, http.StatusForbidden, fmt.Sprintf("该令牌无权使用模型：%s", requestModel))
 				return
 			}
 		}
-		c.Set("id", token.UserId)
-		c.Set("token_id", token.Id)
-		c.Set("token_name", token.Name)
+		c.Set(ctxkey.Id, token.UserId)
+		c.Set(ctxkey.TokenId, token.Id)
+		c.Set(ctxkey.TokenName, token.Name)
 		if len(parts) > 1 {
 			if model.IsAdmin(token.UserId) {
-				c.Set("specific_channel_id", parts[1])
+				c.Set(ctxkey.SpecificChannelId, parts[1])
 			} else {
 				abortWithMessage(c, http.StatusForbidden, "普通用户不支持指定渠道")
 				return

middleware/distributor.go

@@ -3,7 +3,6 @@ package middleware
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/model"
@@ -18,12 +17,12 @@ type ModelRequest struct {
 
 func Distribute() func(c *gin.Context) {
 	return func(c *gin.Context) {
-		userId := c.GetInt("id")
+		userId := c.GetInt(ctxkey.Id)
 		userGroup, _ := model.CacheGetUserGroup(userId)
-		c.Set("group", userGroup)
+		c.Set(ctxkey.Group, userGroup)
 		var requestModel string
 		var channel *model.Channel
-		channelId, ok := c.Get("specific_channel_id")
+		channelId, ok := c.Get(ctxkey.SpecificChannelId)
 		if ok {
 			id, err := strconv.Atoi(channelId.(string))
 			if err != nil {
@@ -40,7 +39,7 @@ func Distribute() func(c *gin.Context) {
 				return
 			}
 		} else {
-			requestModel = c.GetString("request_model")
+			requestModel = c.GetString(ctxkey.RequestModel)
 			var err error
 			channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, requestModel, false)
 			if err != nil {
@@ -59,28 +58,38 @@ func Distribute() func(c *gin.Context) {
 }
 
 func SetupContextForSelectedChannel(c *gin.Context, channel *model.Channel, modelName string) {
-	c.Set("channel", channel.Type)
-	c.Set("channel_id", channel.Id)
-	c.Set("channel_name", channel.Name)
-	c.Set("model_mapping", channel.GetModelMapping())
+	c.Set(ctxkey.Channel, channel.Type)
+	c.Set(ctxkey.ChannelId, channel.Id)
+	c.Set(ctxkey.ChannelName, channel.Name)
+	c.Set(ctxkey.ModelMapping, channel.GetModelMapping())
 	c.Set(ctxkey.OriginalModel, modelName) // for retry
 	c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
-	c.Set("base_url", channel.GetBaseURL())
-	// this is for backward compatibility
-	switch channel.Type {
-	case channeltype.Azure:
-		c.Set(config.KeyAPIVersion, channel.Other)
-	case channeltype.Xunfei:
-		c.Set(config.KeyAPIVersion, channel.Other)
-	case channeltype.Gemini:
-		c.Set(config.KeyAPIVersion, channel.Other)
-	case channeltype.AIProxyLibrary:
-		c.Set(config.KeyLibraryID, channel.Other)
-	case channeltype.Ali:
-		c.Set(config.KeyPlugin, channel.Other)
-	}
+	c.Set(ctxkey.BaseURL, channel.GetBaseURL())
 	cfg, _ := channel.LoadConfig()
-	for k, v := range cfg {
-		c.Set(config.KeyPrefix+k, v)
+	// this is for backward compatibility
+	if channel.Other != nil {
+		switch channel.Type {
+		case channeltype.Azure:
+			if cfg.APIVersion == "" {
+				cfg.APIVersion = *channel.Other
+			}
+		case channeltype.Xunfei:
+			if cfg.APIVersion == "" {
+				cfg.APIVersion = *channel.Other
+			}
+		case channeltype.Gemini:
+			if cfg.APIVersion == "" {
+				cfg.APIVersion = *channel.Other
+			}
+		case channeltype.AIProxyLibrary:
+			if cfg.LibraryID == "" {
+				cfg.LibraryID = *channel.Other
+			}
+		case channeltype.Ali:
+			if cfg.Plugin == "" {
+				cfg.Plugin = *channel.Other
+			}
+		}
 	}
+	c.Set(ctxkey.Config, cfg)
 }

middleware/logger.go

@@ -3,14 +3,14 @@ package middleware
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/helper"
 )
 
 func SetUpLogger(server *gin.Engine) {
 	server.Use(gin.LoggerWithFormatter(func(param gin.LogFormatterParams) string {
 		var requestID string
 		if param.Keys != nil {
-			requestID = param.Keys[logger.RequestIdKey].(string)
+			requestID = param.Keys[helper.RequestIdKey].(string)
 		}
 		return fmt.Sprintf("[GIN] %s | %s | %3d | %13v | %15s | %7s %s\n",
 			param.TimeStamp.Format("2006/01/02 - 15:04:05"),

middleware/request-id.go

@@ -4,16 +4,15 @@ import (
 	"context"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/helper"
-	"github.com/songquanpeng/one-api/common/logger"
 )
 
 func RequestId() func(c *gin.Context) {
 	return func(c *gin.Context) {
 		id := helper.GenRequestID()
-		c.Set(logger.RequestIdKey, id)
-		ctx := context.WithValue(c.Request.Context(), logger.RequestIdKey, id)
+		c.Set(helper.RequestIdKey, id)
+		ctx := context.WithValue(c.Request.Context(), helper.RequestIdKey, id)
 		c.Request = c.Request.WithContext(ctx)
-		c.Header(logger.RequestIdKey, id)
+		c.Header(helper.RequestIdKey, id)
 		c.Next()
 	}
 }

middleware/utils.go

@@ -12,7 +12,7 @@ import (
 func abortWithMessage(c *gin.Context, statusCode int, message string) {
 	c.JSON(statusCode, gin.H{
 		"error": gin.H{
-			"message": helper.MessageWithRequestId(message, c.GetString(logger.RequestIdKey)),
+			"message": helper.MessageWithRequestId(message, c.GetString(helper.RequestIdKey)),
 			"type":    "one_api_error",
 		},
 	})

model/channel.go

@@ -27,7 +27,7 @@ type Channel struct {
 	TestTime           int64   `json:"test_time" gorm:"bigint"`
 	ResponseTime       int     `json:"response_time"` // in milliseconds
 	BaseURL            *string `json:"base_url" gorm:"column:base_url;default:''"`
-	Other              string  `json:"other"`   // DEPRECATED: please save config to field Config
+	Other              *string `json:"other"`   // DEPRECATED: please save config to field Config
 	Balance            float64 `json:"balance"` // in USD
 	BalanceUpdatedTime int64   `json:"balance_updated_time" gorm:"bigint"`
 	Models             string  `json:"models"`
@@ -38,6 +38,16 @@ type Channel struct {
 	Config             string  `json:"config"`
 }
 
+type ChannelConfig struct {
+	Region     string `json:"region,omitempty"`
+	SK         string `json:"sk,omitempty"`
+	AK         string `json:"ak,omitempty"`
+	UserID     string `json:"user_id,omitempty"`
+	APIVersion string `json:"api_version,omitempty"`
+	LibraryID  string `json:"library_id,omitempty"`
+	Plugin     string `json:"plugin,omitempty"`
+}
+
 func GetAllChannels(startIdx int, num int, scope string) ([]*Channel, error) {
 	var channels []*Channel
 	var err error
@@ -161,14 +171,14 @@ func (channel *Channel) Delete() error {
 	return err
 }
 
-func (channel *Channel) LoadConfig() (map[string]string, error) {
+func (channel *Channel) LoadConfig() (ChannelConfig, error) {
+	var cfg ChannelConfig
 	if channel.Config == "" {
-		return nil, nil
+		return cfg, nil
 	}
-	cfg := make(map[string]string)
 	err := json.Unmarshal([]byte(channel.Config), &cfg)
 	if err != nil {
-		return nil, err
+		return cfg, err
 	}
 	return cfg, nil
 }

model/user.go

@@ -6,6 +6,7 @@ import (
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/blacklist"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/common/random"
 	"gorm.io/gorm"
@@ -140,6 +141,22 @@ func (user *User) Insert(inviterId int) error {
 			RecordLog(inviterId, LogTypeSystem, fmt.Sprintf("邀请用户赠送 %s", common.LogQuota(config.QuotaForInviter)))
 		}
 	}
+	// create default token
+	cleanToken := Token{
+		UserId:         user.Id,
+		Name:           "default",
+		Key:            random.GenerateKey(),
+		CreatedTime:    helper.GetTimestamp(),
+		AccessedTime:   helper.GetTimestamp(),
+		ExpiredTime:    -1,
+		RemainQuota:    -1,
+		UnlimitedQuota: true,
+	}
+	result.Error = cleanToken.Insert()
+	if result.Error != nil {
+		// do not block
+		logger.SysError(fmt.Sprintf("create default token for user %d failed: %s", user.Id, result.Error.Error()))
+	}
 	return nil
 }
 

relay/adaptor.go

@@ -7,6 +7,10 @@ import (
 	"github.com/songquanpeng/one-api/relay/adaptor/anthropic"
 	"github.com/songquanpeng/one-api/relay/adaptor/aws"
 	"github.com/songquanpeng/one-api/relay/adaptor/baidu"
+	"github.com/songquanpeng/one-api/relay/adaptor/cloudflare"
+	"github.com/songquanpeng/one-api/relay/adaptor/cohere"
+	"github.com/songquanpeng/one-api/relay/adaptor/coze"
+	"github.com/songquanpeng/one-api/relay/adaptor/deepl"
 	"github.com/songquanpeng/one-api/relay/adaptor/gemini"
 	"github.com/songquanpeng/one-api/relay/adaptor/ollama"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
@@ -43,6 +47,14 @@ func GetAdaptor(apiType int) adaptor.Adaptor {
 		return &zhipu.Adaptor{}
 	case apitype.Ollama:
 		return &ollama.Adaptor{}
+	case apitype.Coze:
+		return &coze.Adaptor{}
+	case apitype.Cohere:
+		return &cohere.Adaptor{}
+	case apitype.Cloudflare:
+		return &cloudflare.Adaptor{}
+	case apitype.DeepL:
+		return &deepl.Adaptor{}
 	}
 	return nil
 }

relay/adaptor/aiproxy/adaptor.go

@@ -4,7 +4,6 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/relay/adaptor"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -13,10 +12,11 @@ import (
 )
 
 type Adaptor struct {
+	meta *meta.Meta
 }
 
 func (a *Adaptor) Init(meta *meta.Meta) {
-
+	a.meta = meta
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
@@ -34,7 +34,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 		return nil, errors.New("request is nil")
 	}
 	aiProxyLibraryRequest := ConvertRequest(*request)
-	aiProxyLibraryRequest.LibraryId = c.GetString(config.KeyLibraryID)
+	aiProxyLibraryRequest.LibraryId = a.meta.Config.LibraryID
 	return aiProxyLibraryRequest, nil
 }
 

relay/adaptor/ali/adaptor.go

@@ -4,7 +4,6 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/relay/adaptor"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -16,10 +15,11 @@ import (
 // https://help.aliyun.com/zh/dashscope/developer-reference/api-details
 
 type Adaptor struct {
+	meta *meta.Meta
 }
 
 func (a *Adaptor) Init(meta *meta.Meta) {
-
+	a.meta = meta
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
@@ -47,8 +47,8 @@ func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *me
 	if meta.Mode == relaymode.ImagesGenerations {
 		req.Header.Set("X-DashScope-Async", "enable")
 	}
-	if c.GetString(config.KeyPlugin) != "" {
-		req.Header.Set("X-DashScope-Plugin", c.GetString(config.KeyPlugin))
+	if a.meta.Config.Plugin != "" {
+		req.Header.Set("X-DashScope-Plugin", a.meta.Config.Plugin)
 	}
 	return nil
 }

relay/adaptor/anthropic/main.go

@@ -4,6 +4,10 @@ import (
 	"bufio"
 	"encoding/json"
 	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/helper"
@@ -11,9 +15,6 @@ import (
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/model"
-	"io"
-	"net/http"
-	"strings"
 )
 
 func stopReasonClaude2OpenAI(reason *string) string {
@@ -176,10 +177,10 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 			if len(data) < 6 {
 				continue
 			}
-			if !strings.HasPrefix(data, "data: ") {
+			if !strings.HasPrefix(data, "data:") {
 				continue
 			}
-			data = strings.TrimPrefix(data, "data: ")
+			data = strings.TrimPrefix(data, "data:")
 			dataChan <- data
 		}
 		stopChan <- true
@@ -192,7 +193,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 		select {
 		case data := <-dataChan:
 			// some implementations may add \r at the end of data
-			data = strings.TrimSuffix(data, "\r")
+			data = strings.TrimSpace(data)
 			var claudeResponse StreamResponse
 			err := json.Unmarshal([]byte(data), &claudeResponse)
 			if err != nil {

relay/adaptor/aws/adapter.go

@@ -1,6 +1,9 @@
 package aws
 
 import (
+	"github.com/aws/aws-sdk-go-v2/aws"
+	"github.com/aws/aws-sdk-go-v2/credentials"
+	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime"
 	"github.com/songquanpeng/one-api/common/ctxkey"
 	"io"
 	"net/http"
@@ -16,10 +19,16 @@ import (
 var _ adaptor.Adaptor = new(Adaptor)
 
 type Adaptor struct {
+	meta      *meta.Meta
+	awsClient *bedrockruntime.Client
 }
 
 func (a *Adaptor) Init(meta *meta.Meta) {
-
+	a.meta = meta
+	a.awsClient = bedrockruntime.New(bedrockruntime.Options{
+		Region:      meta.Config.Region,
+		Credentials: aws.NewCredentialsCache(credentials.NewStaticCredentialsProvider(meta.Config.AK, meta.Config.SK, "")),
+	})
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
@@ -54,9 +63,9 @@ func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Read
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
 	if meta.IsStream {
-		err, usage = StreamHandler(c, resp)
+		err, usage = StreamHandler(c, a.awsClient)
 	} else {
-		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+		err, usage = Handler(c, a.awsClient, meta.ActualModelName)
 	}
 	return
 }
@@ -65,7 +74,6 @@ func (a *Adaptor) GetModelList() (models []string) {
 	for n := range awsModelIDMap {
 		models = append(models, n)
 	}
-
 	return
 }
 

relay/adaptor/aws/main.go

@@ -5,13 +5,11 @@ import (
 	"bytes"
 	"encoding/json"
 	"fmt"
-	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/ctxkey"
 	"io"
 	"net/http"
 
 	"github.com/aws/aws-sdk-go-v2/aws"
-	"github.com/aws/aws-sdk-go-v2/credentials"
 	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime"
 	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime/types"
 	"github.com/gin-gonic/gin"
@@ -24,18 +22,6 @@ import (
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
 )
 
-func newAwsClient(c *gin.Context) (*bedrockruntime.Client, error) {
-	ak := c.GetString(config.KeyAK)
-	sk := c.GetString(config.KeySK)
-	region := c.GetString(config.KeyRegion)
-	client := bedrockruntime.New(bedrockruntime.Options{
-		Region:      region,
-		Credentials: aws.NewCredentialsCache(credentials.NewStaticCredentialsProvider(ak, sk, "")),
-	})
-
-	return client, nil
-}
-
 func wrapErr(err error) *relaymodel.ErrorWithStatusCode {
 	return &relaymodel.ErrorWithStatusCode{
 		StatusCode: http.StatusInternalServerError,
@@ -63,12 +49,7 @@ func awsModelID(requestModel string) (string, error) {
 	return "", errors.Errorf("model %s not found", requestModel)
 }
 
-func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
-	awsCli, err := newAwsClient(c)
-	if err != nil {
-		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
-	}
-
+func Handler(c *gin.Context, awsCli *bedrockruntime.Client, modelName string) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
 	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return wrapErr(errors.Wrap(err, "awsModelID")), nil
@@ -121,13 +102,8 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 	return nil, &usage
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
+func StreamHandler(c *gin.Context, awsCli *bedrockruntime.Client) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
 	createdTime := helper.GetTimestamp()
-	awsCli, err := newAwsClient(c)
-	if err != nil {
-		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
-	}
-
 	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return wrapErr(errors.Wrap(err, "awsModelID")), nil

relay/adaptor/azure/helper.go

@@ -1,15 +0,0 @@
-package azure
-
-import (
-	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/common/config"
-)
-
-func GetAPIVersion(c *gin.Context) string {
-	query := c.Request.URL.Query()
-	apiVersion := query.Get("api-version")
-	if apiVersion == "" {
-		apiVersion = c.GetString(config.KeyAPIVersion)
-	}
-	return apiVersion
-}

relay/adaptor/baidu/main.go

@@ -7,9 +7,9 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/client"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
-	"github.com/songquanpeng/one-api/relay/client"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/model"
 	"io"

relay/adaptor/cloudflare/adaptor.go

@@ -0,0 +1,66 @@
+package cloudflare
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/adaptor"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+type Adaptor struct {
+	meta *meta.Meta
+}
+
+// ConvertImageRequest implements adaptor.Adaptor.
+func (*Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	return nil, errors.New("not implemented")
+}
+
+// ConvertImageRequest implements adaptor.Adaptor.
+
+func (a *Adaptor) Init(meta *meta.Meta) {
+	a.meta = meta
+}
+
+func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
+	return fmt.Sprintf("%s/client/v4/accounts/%s/ai/run/%s", meta.BaseURL, meta.Config.UserID, meta.ActualModelName), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *meta.Meta) error {
+	adaptor.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return ConvertRequest(*request), nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Reader) (*http.Response, error) {
+	return adaptor.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "cloudflare"
+}

relay/adaptor/cloudflare/constant.go

@@ -0,0 +1,36 @@
+package cloudflare
+
+var ModelList = []string{
+	"@cf/meta/llama-2-7b-chat-fp16",
+	"@cf/meta/llama-2-7b-chat-int8",
+	"@cf/mistral/mistral-7b-instruct-v0.1",
+	"@hf/thebloke/deepseek-coder-6.7b-base-awq",
+	"@hf/thebloke/deepseek-coder-6.7b-instruct-awq",
+	"@cf/deepseek-ai/deepseek-math-7b-base",
+	"@cf/deepseek-ai/deepseek-math-7b-instruct",
+	"@cf/thebloke/discolm-german-7b-v1-awq",
+	"@cf/tiiuae/falcon-7b-instruct",
+	"@cf/google/gemma-2b-it-lora",
+	"@hf/google/gemma-7b-it",
+	"@cf/google/gemma-7b-it-lora",
+	"@hf/nousresearch/hermes-2-pro-mistral-7b",
+	"@hf/thebloke/llama-2-13b-chat-awq",
+	"@cf/meta-llama/llama-2-7b-chat-hf-lora",
+	"@cf/meta/llama-3-8b-instruct",
+	"@hf/thebloke/llamaguard-7b-awq",
+	"@hf/thebloke/mistral-7b-instruct-v0.1-awq",
+	"@hf/mistralai/mistral-7b-instruct-v0.2",
+	"@cf/mistral/mistral-7b-instruct-v0.2-lora",
+	"@hf/thebloke/neural-chat-7b-v3-1-awq",
+	"@cf/openchat/openchat-3.5-0106",
+	"@hf/thebloke/openhermes-2.5-mistral-7b-awq",
+	"@cf/microsoft/phi-2",
+	"@cf/qwen/qwen1.5-0.5b-chat",
+	"@cf/qwen/qwen1.5-1.8b-chat",
+	"@cf/qwen/qwen1.5-14b-chat-awq",
+	"@cf/qwen/qwen1.5-7b-chat-awq",
+	"@cf/defog/sqlcoder-7b-2",
+	"@hf/nexusflow/starling-lm-7b-beta",
+	"@cf/tinyllama/tinyllama-1.1b-chat-v1.0",
+	"@hf/thebloke/zephyr-7b-beta-awq",
+}

relay/adaptor/cloudflare/main.go

@@ -0,0 +1,158 @@
+package cloudflare
+
+import (
+	"bufio"
+	"bytes"
+	"encoding/json"
+	"io"
+	"net/http"
+	"strings"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
+    var promptBuilder strings.Builder
+    for _, message := range textRequest.Messages {
+        promptBuilder.WriteString(message.StringContent())
+        promptBuilder.WriteString("\n")  // 添加换行符来分隔每个消息
+    }
+
+    return &Request{
+        MaxTokens:   textRequest.MaxTokens,
+        Prompt:      promptBuilder.String(),
+        Stream:      textRequest.Stream,
+        Temperature: textRequest.Temperature,
+    }
+}
+
+
+func ResponseCloudflare2OpenAI(cloudflareResponse *Response) *openai.TextResponse {
+	choice := openai.TextResponseChoice{
+		Index: 0,
+		Message: model.Message{
+			Role:    "assistant",
+			Content: cloudflareResponse.Result.Response,
+		},
+		FinishReason: "stop",
+	}
+	fullTextResponse := openai.TextResponse{
+		Object:  "chat.completion",
+		Created: helper.GetTimestamp(),
+		Choices: []openai.TextResponseChoice{choice},
+	}
+	return &fullTextResponse
+}
+
+func StreamResponseCloudflare2OpenAI(cloudflareResponse *StreamResponse) *openai.ChatCompletionsStreamResponse {
+	var choice openai.ChatCompletionsStreamResponseChoice
+	choice.Delta.Content = cloudflareResponse.Response
+	choice.Delta.Role = "assistant"
+	openaiResponse := openai.ChatCompletionsStreamResponse{
+		Object:  "chat.completion.chunk",
+		Choices: []openai.ChatCompletionsStreamResponseChoice{choice},
+		Created: helper.GetTimestamp(),
+	}
+	return &openaiResponse
+}
+
+func StreamHandler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
+	scanner := bufio.NewScanner(resp.Body)
+	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
+		if atEOF && len(data) == 0 {
+			return 0, nil, nil
+		}
+		if i := bytes.IndexByte(data, '\n'); i >= 0 {
+			return i + 1, data[0:i], nil
+		}
+		if atEOF {
+			return len(data), data, nil
+		}
+		return 0, nil, nil
+	})
+
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
+	go func() {
+		for scanner.Scan() {
+			data := scanner.Text()
+			if len(data) < len("data: ") {
+				continue
+			}
+			data = strings.TrimPrefix(data, "data: ")
+			dataChan <- data
+		}
+		stopChan <- true
+	}()
+	common.SetEventStreamHeaders(c)
+	id := helper.GetResponseID(c)
+	responseModel := c.GetString("original_model")
+	var responseText string
+	c.Stream(func(w io.Writer) bool {
+		select {
+		case data := <-dataChan:
+			// some implementations may add \r at the end of data
+			data = strings.TrimSuffix(data, "\r")
+			var cloudflareResponse StreamResponse
+			err := json.Unmarshal([]byte(data), &cloudflareResponse)
+			if err != nil {
+				logger.SysError("error unmarshalling stream response: " + err.Error())
+				return true
+			}
+			response := StreamResponseCloudflare2OpenAI(&cloudflareResponse)
+			if response == nil {
+				return true
+			}
+			responseText += cloudflareResponse.Response
+			response.Id = id
+			response.Model = responseModel
+			jsonStr, err := json.Marshal(response)
+			if err != nil {
+				logger.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonStr)})
+			return true
+		case <-stopChan:
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+	})
+	_ = resp.Body.Close()
+	usage := openai.ResponseText2Usage(responseText, responseModel, promptTokens)
+	return nil, usage
+}
+
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	var cloudflareResponse Response
+	err = json.Unmarshal(responseBody, &cloudflareResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	fullTextResponse := ResponseCloudflare2OpenAI(&cloudflareResponse)
+	fullTextResponse.Model = modelName
+	usage := openai.ResponseText2Usage(cloudflareResponse.Result.Response, modelName, promptTokens)
+	fullTextResponse.Usage = *usage
+	fullTextResponse.Id = helper.GetResponseID(c)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, usage
+}

relay/adaptor/cloudflare/model.go

@@ -0,0 +1,25 @@
+package cloudflare
+
+type Request struct {
+	Lora        string  `json:"lora,omitempty"`
+	MaxTokens   int     `json:"max_tokens,omitempty"`
+	Prompt      string  `json:"prompt,omitempty"`
+	Raw         bool    `json:"raw,omitempty"`
+	Stream      bool    `json:"stream,omitempty"`
+	Temperature float64 `json:"temperature,omitempty"`
+}
+
+type Result struct {
+	Response string `json:"response"`
+}
+
+type Response struct {
+	Result   Result   `json:"result"`
+	Success  bool     `json:"success"`
+	Errors   []string `json:"errors"`
+	Messages []string `json:"messages"`
+}
+
+type StreamResponse struct {
+	Response string `json:"response"`
+}

relay/adaptor/cohere/adaptor.go

@@ -0,0 +1,64 @@
+package cohere
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/adaptor"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+type Adaptor struct{}
+
+// ConvertImageRequest implements adaptor.Adaptor.
+func (*Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	return nil, errors.New("not implemented")
+}
+
+// ConvertImageRequest implements adaptor.Adaptor.
+
+func (a *Adaptor) Init(meta *meta.Meta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
+	return fmt.Sprintf("%s/v1/chat", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *meta.Meta) error {
+	adaptor.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return ConvertRequest(*request), nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Reader) (*http.Response, error) {
+	return adaptor.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "Cohere"
+}

relay/adaptor/cohere/constant.go

@@ -0,0 +1,14 @@
+package cohere
+
+var ModelList = []string{
+	"command", "command-nightly",
+	"command-light", "command-light-nightly",
+	"command-r", "command-r-plus",
+}
+
+func init() {
+	num := len(ModelList)
+	for i := 0; i < num; i++ {
+		ModelList = append(ModelList, ModelList[i]+"-internet")
+	}
+}

relay/adaptor/cohere/main.go

@@ -0,0 +1,241 @@
+package cohere
+
+import (
+	"bufio"
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+var (
+	WebSearchConnector = Connector{ID: "web-search"}
+)
+
+func stopReasonCohere2OpenAI(reason *string) string {
+	if reason == nil {
+		return ""
+	}
+	switch *reason {
+	case "COMPLETE":
+		return "stop"
+	default:
+		return *reason
+	}
+}
+
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
+	cohereRequest := Request{
+		Model:            textRequest.Model,
+		Message:          "",
+		MaxTokens:        textRequest.MaxTokens,
+		Temperature:      textRequest.Temperature,
+		P:                textRequest.TopP,
+		K:                textRequest.TopK,
+		Stream:           textRequest.Stream,
+		FrequencyPenalty: textRequest.FrequencyPenalty,
+		PresencePenalty:  textRequest.FrequencyPenalty,
+		Seed:             int(textRequest.Seed),
+	}
+	if cohereRequest.Model == "" {
+		cohereRequest.Model = "command-r"
+	}
+	if strings.HasSuffix(cohereRequest.Model, "-internet") {
+		cohereRequest.Model = strings.TrimSuffix(cohereRequest.Model, "-internet")
+		cohereRequest.Connectors = append(cohereRequest.Connectors, WebSearchConnector)
+	}
+	for _, message := range textRequest.Messages {
+		if message.Role == "user" {
+			cohereRequest.Message = message.Content.(string)
+		} else {
+			var role string
+			if message.Role == "assistant" {
+				role = "CHATBOT"
+			} else if message.Role == "system" {
+				role = "SYSTEM"
+			} else {
+				role = "USER"
+			}
+			cohereRequest.ChatHistory = append(cohereRequest.ChatHistory, ChatMessage{
+				Role:    role,
+				Message: message.Content.(string),
+			})
+		}
+	}
+	return &cohereRequest
+}
+
+func StreamResponseCohere2OpenAI(cohereResponse *StreamResponse) (*openai.ChatCompletionsStreamResponse, *Response) {
+	var response *Response
+	var responseText string
+	var finishReason string
+
+	switch cohereResponse.EventType {
+	case "stream-start":
+		return nil, nil
+	case "text-generation":
+		responseText += cohereResponse.Text
+	case "stream-end":
+		usage := cohereResponse.Response.Meta.Tokens
+		response = &Response{
+			Meta: Meta{
+				Tokens: Usage{
+					InputTokens:  usage.InputTokens,
+					OutputTokens: usage.OutputTokens,
+				},
+			},
+		}
+		finishReason = *cohereResponse.Response.FinishReason
+	default:
+		return nil, nil
+	}
+
+	var choice openai.ChatCompletionsStreamResponseChoice
+	choice.Delta.Content = responseText
+	choice.Delta.Role = "assistant"
+	if finishReason != "" {
+		choice.FinishReason = &finishReason
+	}
+	var openaiResponse openai.ChatCompletionsStreamResponse
+	openaiResponse.Object = "chat.completion.chunk"
+	openaiResponse.Choices = []openai.ChatCompletionsStreamResponseChoice{choice}
+	return &openaiResponse, response
+}
+
+func ResponseCohere2OpenAI(cohereResponse *Response) *openai.TextResponse {
+	choice := openai.TextResponseChoice{
+		Index: 0,
+		Message: model.Message{
+			Role:    "assistant",
+			Content: cohereResponse.Text,
+			Name:    nil,
+		},
+		FinishReason: stopReasonCohere2OpenAI(cohereResponse.FinishReason),
+	}
+	fullTextResponse := openai.TextResponse{
+		Id:      fmt.Sprintf("chatcmpl-%s", cohereResponse.ResponseID),
+		Model:   "model",
+		Object:  "chat.completion",
+		Created: helper.GetTimestamp(),
+		Choices: []openai.TextResponseChoice{choice},
+	}
+	return &fullTextResponse
+}
+
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	createdTime := helper.GetTimestamp()
+	scanner := bufio.NewScanner(resp.Body)
+	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
+		if atEOF && len(data) == 0 {
+			return 0, nil, nil
+		}
+		if i := bytes.IndexByte(data, '\n'); i >= 0 {
+			return i + 1, data[0:i], nil
+		}
+		if atEOF {
+			return len(data), data, nil
+		}
+		return 0, nil, nil
+	})
+
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
+	go func() {
+		for scanner.Scan() {
+			data := scanner.Text()
+			dataChan <- data
+		}
+		stopChan <- true
+	}()
+	common.SetEventStreamHeaders(c)
+	var usage model.Usage
+	c.Stream(func(w io.Writer) bool {
+		select {
+		case data := <-dataChan:
+			// some implementations may add \r at the end of data
+			data = strings.TrimSuffix(data, "\r")
+			var cohereResponse StreamResponse
+			err := json.Unmarshal([]byte(data), &cohereResponse)
+			if err != nil {
+				logger.SysError("error unmarshalling stream response: " + err.Error())
+				return true
+			}
+			response, meta := StreamResponseCohere2OpenAI(&cohereResponse)
+			if meta != nil {
+				usage.PromptTokens += meta.Meta.Tokens.InputTokens
+				usage.CompletionTokens += meta.Meta.Tokens.OutputTokens
+				return true
+			}
+			if response == nil {
+				return true
+			}
+			response.Id = fmt.Sprintf("chatcmpl-%d", createdTime)
+			response.Model = c.GetString("original_model")
+			response.Created = createdTime
+			jsonStr, err := json.Marshal(response)
+			if err != nil {
+				logger.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonStr)})
+			return true
+		case <-stopChan:
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+	})
+	_ = resp.Body.Close()
+	return nil, &usage
+}
+
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	var cohereResponse Response
+	err = json.Unmarshal(responseBody, &cohereResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	if cohereResponse.ResponseID == "" {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: cohereResponse.Message,
+				Type:    cohereResponse.Message,
+				Param:   "",
+				Code:    resp.StatusCode,
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+	fullTextResponse := ResponseCohere2OpenAI(&cohereResponse)
+	fullTextResponse.Model = modelName
+	usage := model.Usage{
+		PromptTokens:     cohereResponse.Meta.Tokens.InputTokens,
+		CompletionTokens: cohereResponse.Meta.Tokens.OutputTokens,
+		TotalTokens:      cohereResponse.Meta.Tokens.InputTokens + cohereResponse.Meta.Tokens.OutputTokens,
+	}
+	fullTextResponse.Usage = usage
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &usage
+}

relay/adaptor/cohere/model.go

@@ -0,0 +1,147 @@
+package cohere
+
+type Request struct {
+	Message          string        `json:"message" required:"true"`
+	Model            string        `json:"model,omitempty"`  // 默认值为"command-r"
+	Stream           bool          `json:"stream,omitempty"` // 默认值为false
+	Preamble         string        `json:"preamble,omitempty"`
+	ChatHistory      []ChatMessage `json:"chat_history,omitempty"`
+	ConversationID   string        `json:"conversation_id,omitempty"`
+	PromptTruncation string        `json:"prompt_truncation,omitempty"` // 默认值为"AUTO"
+	Connectors       []Connector   `json:"connectors,omitempty"`
+	Documents        []Document    `json:"documents,omitempty"`
+	Temperature      float64       `json:"temperature,omitempty"` // 默认值为0.3
+	MaxTokens        int           `json:"max_tokens,omitempty"`
+	MaxInputTokens   int           `json:"max_input_tokens,omitempty"`
+	K                int           `json:"k,omitempty"` // 默认值为0
+	P                float64       `json:"p,omitempty"` // 默认值为0.75
+	Seed             int           `json:"seed,omitempty"`
+	StopSequences    []string      `json:"stop_sequences,omitempty"`
+	FrequencyPenalty float64       `json:"frequency_penalty,omitempty"` // 默认值为0.0
+	PresencePenalty  float64       `json:"presence_penalty,omitempty"`  // 默认值为0.0
+	Tools            []Tool        `json:"tools,omitempty"`
+	ToolResults      []ToolResult  `json:"tool_results,omitempty"`
+}
+
+type ChatMessage struct {
+	Role    string `json:"role" required:"true"`
+	Message string `json:"message" required:"true"`
+}
+
+type Tool struct {
+	Name                 string                   `json:"name" required:"true"`
+	Description          string                   `json:"description" required:"true"`
+	ParameterDefinitions map[string]ParameterSpec `json:"parameter_definitions"`
+}
+
+type ParameterSpec struct {
+	Description string `json:"description"`
+	Type        string `json:"type" required:"true"`
+	Required    bool   `json:"required"`
+}
+
+type ToolResult struct {
+	Call    ToolCall                 `json:"call"`
+	Outputs []map[string]interface{} `json:"outputs"`
+}
+
+type ToolCall struct {
+	Name       string                 `json:"name" required:"true"`
+	Parameters map[string]interface{} `json:"parameters" required:"true"`
+}
+
+type StreamResponse struct {
+	IsFinished    bool            `json:"is_finished"`
+	EventType     string          `json:"event_type"`
+	GenerationID  string          `json:"generation_id,omitempty"`
+	SearchQueries []*SearchQuery  `json:"search_queries,omitempty"`
+	SearchResults []*SearchResult `json:"search_results,omitempty"`
+	Documents     []*Document     `json:"documents,omitempty"`
+	Text          string          `json:"text,omitempty"`
+	Citations     []*Citation     `json:"citations,omitempty"`
+	Response      *Response       `json:"response,omitempty"`
+	FinishReason  string          `json:"finish_reason,omitempty"`
+}
+
+type SearchQuery struct {
+	Text         string `json:"text"`
+	GenerationID string `json:"generation_id"`
+}
+
+type SearchResult struct {
+	SearchQuery *SearchQuery `json:"search_query"`
+	DocumentIDs []string     `json:"document_ids"`
+	Connector   *Connector   `json:"connector"`
+}
+
+type Connector struct {
+	ID string `json:"id"`
+}
+
+type Document struct {
+	ID        string `json:"id"`
+	Snippet   string `json:"snippet"`
+	Timestamp string `json:"timestamp"`
+	Title     string `json:"title"`
+	URL       string `json:"url"`
+}
+
+type Citation struct {
+	Start       int      `json:"start"`
+	End         int      `json:"end"`
+	Text        string   `json:"text"`
+	DocumentIDs []string `json:"document_ids"`
+}
+
+type Response struct {
+	ResponseID    string          `json:"response_id"`
+	Text          string          `json:"text"`
+	GenerationID  string          `json:"generation_id"`
+	ChatHistory   []*Message      `json:"chat_history"`
+	FinishReason  *string         `json:"finish_reason"`
+	Meta          Meta            `json:"meta"`
+	Citations     []*Citation     `json:"citations"`
+	Documents     []*Document     `json:"documents"`
+	SearchResults []*SearchResult `json:"search_results"`
+	SearchQueries []*SearchQuery  `json:"search_queries"`
+	Message       string          `json:"message"`
+}
+
+type Message struct {
+	Role    string `json:"role"`
+	Message string `json:"message"`
+}
+
+type Version struct {
+	Version string `json:"version"`
+}
+
+type Units struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
+}
+
+type ChatEntry struct {
+	Role    string `json:"role"`
+	Message string `json:"message"`
+}
+
+type Meta struct {
+	APIVersion  APIVersion  `json:"api_version"`
+	BilledUnits BilledUnits `json:"billed_units"`
+	Tokens      Usage       `json:"tokens"`
+}
+
+type APIVersion struct {
+	Version string `json:"version"`
+}
+
+type BilledUnits struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
+}
+
+type Usage struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
+}

relay/adaptor/common.go

@@ -4,7 +4,7 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/relay/client"
+	"github.com/songquanpeng/one-api/common/client"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"io"
 	"net/http"

relay/adaptor/coze/adaptor.go

@@ -0,0 +1,75 @@
+package coze
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/adaptor"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/model"
+	"io"
+	"net/http"
+)
+
+type Adaptor struct {
+	meta *meta.Meta
+}
+
+func (a *Adaptor) Init(meta *meta.Meta) {
+	a.meta = meta
+}
+
+func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
+	return fmt.Sprintf("%s/open_api/v2/chat", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *meta.Meta) error {
+	adaptor.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	request.User = a.meta.Config.UserID
+	return ConvertRequest(*request), nil
+}
+
+func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return request, nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Reader) (*http.Response, error) {
+	return adaptor.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	var responseText *string
+	if meta.IsStream {
+		err, responseText = StreamHandler(c, resp)
+	} else {
+		err, responseText = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	if responseText != nil {
+		usage = openai.ResponseText2Usage(*responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		usage = &model.Usage{}
+	}
+	usage.PromptTokens = meta.PromptTokens
+	usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "coze"
+}

relay/adaptor/coze/constant/contenttype/define.go

@@ -0,0 +1,5 @@
+package contenttype
+
+const (
+	Text = "text"
+)

relay/adaptor/coze/constant/event/define.go

@@ -0,0 +1,7 @@
+package event
+
+const (
+	Message = "message"
+	Done    = "done"
+	Error   = "error"
+)

relay/adaptor/coze/constant/messagetype/define.go

@@ -0,0 +1,6 @@
+package messagetype
+
+const (
+	Answer   = "answer"
+	FollowUp = "follow_up"
+)

relay/adaptor/coze/constants.go

@@ -0,0 +1,3 @@
+package coze
+
+var ModelList = []string{}

relay/adaptor/coze/helper.go

@@ -0,0 +1,10 @@
+package coze
+
+import "github.com/songquanpeng/one-api/relay/adaptor/coze/constant/event"
+
+func event2StopReason(e *string) string {
+	if e == nil || *e == event.Message {
+		return ""
+	}
+	return "stop"
+}

relay/adaptor/coze/main.go

@@ -0,0 +1,215 @@
+package coze
+
+import (
+	"bufio"
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/conv"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/adaptor/coze/constant/messagetype"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+	"io"
+	"net/http"
+	"strings"
+)
+
+// https://www.coze.com/open
+
+func stopReasonCoze2OpenAI(reason *string) string {
+	if reason == nil {
+		return ""
+	}
+	switch *reason {
+	case "end_turn":
+		return "stop"
+	case "stop_sequence":
+		return "stop"
+	case "max_tokens":
+		return "length"
+	default:
+		return *reason
+	}
+}
+
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
+	cozeRequest := Request{
+		Stream: textRequest.Stream,
+		User:   textRequest.User,
+		BotId:  strings.TrimPrefix(textRequest.Model, "bot-"),
+	}
+	for i, message := range textRequest.Messages {
+		if i == len(textRequest.Messages)-1 {
+			cozeRequest.Query = message.StringContent()
+			continue
+		}
+		cozeMessage := Message{
+			Role:    message.Role,
+			Content: message.StringContent(),
+		}
+		cozeRequest.ChatHistory = append(cozeRequest.ChatHistory, cozeMessage)
+	}
+	return &cozeRequest
+}
+
+func StreamResponseCoze2OpenAI(cozeResponse *StreamResponse) (*openai.ChatCompletionsStreamResponse, *Response) {
+	var response *Response
+	var stopReason string
+	var choice openai.ChatCompletionsStreamResponseChoice
+
+	if cozeResponse.Message != nil {
+		if cozeResponse.Message.Type != messagetype.Answer {
+			return nil, nil
+		}
+		choice.Delta.Content = cozeResponse.Message.Content
+	}
+	choice.Delta.Role = "assistant"
+	finishReason := stopReasonCoze2OpenAI(&stopReason)
+	if finishReason != "null" {
+		choice.FinishReason = &finishReason
+	}
+	var openaiResponse openai.ChatCompletionsStreamResponse
+	openaiResponse.Object = "chat.completion.chunk"
+	openaiResponse.Choices = []openai.ChatCompletionsStreamResponseChoice{choice}
+	openaiResponse.Id = cozeResponse.ConversationId
+	return &openaiResponse, response
+}
+
+func ResponseCoze2OpenAI(cozeResponse *Response) *openai.TextResponse {
+	var responseText string
+	for _, message := range cozeResponse.Messages {
+		if message.Type == messagetype.Answer {
+			responseText = message.Content
+			break
+		}
+	}
+	choice := openai.TextResponseChoice{
+		Index: 0,
+		Message: model.Message{
+			Role:    "assistant",
+			Content: responseText,
+			Name:    nil,
+		},
+		FinishReason: "stop",
+	}
+	fullTextResponse := openai.TextResponse{
+		Id:      fmt.Sprintf("chatcmpl-%s", cozeResponse.ConversationId),
+		Model:   "coze-bot",
+		Object:  "chat.completion",
+		Created: helper.GetTimestamp(),
+		Choices: []openai.TextResponseChoice{choice},
+	}
+	return &fullTextResponse
+}
+
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *string) {
+	var responseText string
+	createdTime := helper.GetTimestamp()
+	scanner := bufio.NewScanner(resp.Body)
+	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
+		if atEOF && len(data) == 0 {
+			return 0, nil, nil
+		}
+		if i := strings.Index(string(data), "\n"); i >= 0 {
+			return i + 1, data[0:i], nil
+		}
+		if atEOF {
+			return len(data), data, nil
+		}
+		return 0, nil, nil
+	})
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
+	go func() {
+		for scanner.Scan() {
+			data := scanner.Text()
+			if len(data) < 5 {
+				continue
+			}
+			if !strings.HasPrefix(data, "data:") {
+				continue
+			}
+			data = strings.TrimPrefix(data, "data:")
+			dataChan <- data
+		}
+		stopChan <- true
+	}()
+	common.SetEventStreamHeaders(c)
+	var modelName string
+	c.Stream(func(w io.Writer) bool {
+		select {
+		case data := <-dataChan:
+			// some implementations may add \r at the end of data
+			data = strings.TrimSuffix(data, "\r")
+			var cozeResponse StreamResponse
+			err := json.Unmarshal([]byte(data), &cozeResponse)
+			if err != nil {
+				logger.SysError("error unmarshalling stream response: " + err.Error())
+				return true
+			}
+			response, _ := StreamResponseCoze2OpenAI(&cozeResponse)
+			if response == nil {
+				return true
+			}
+			for _, choice := range response.Choices {
+				responseText += conv.AsString(choice.Delta.Content)
+			}
+			response.Model = modelName
+			response.Created = createdTime
+			jsonStr, err := json.Marshal(response)
+			if err != nil {
+				logger.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonStr)})
+			return true
+		case <-stopChan:
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+	})
+	_ = resp.Body.Close()
+	return nil, &responseText
+}
+
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *string) {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	var cozeResponse Response
+	err = json.Unmarshal(responseBody, &cozeResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	if cozeResponse.Code != 0 {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: cozeResponse.Msg,
+				Code:    cozeResponse.Code,
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+	fullTextResponse := ResponseCoze2OpenAI(&cozeResponse)
+	fullTextResponse.Model = modelName
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	var responseText string
+	if len(fullTextResponse.Choices) > 0 {
+		responseText = fullTextResponse.Choices[0].Message.StringContent()
+	}
+	return nil, &responseText
+}

relay/adaptor/coze/model.go

@@ -0,0 +1,38 @@
+package coze
+
+type Message struct {
+	Role        string `json:"role"`
+	Type        string `json:"type"`
+	Content     string `json:"content"`
+	ContentType string `json:"content_type"`
+}
+
+type ErrorInformation struct {
+	Code int    `json:"code"`
+	Msg  string `json:"msg"`
+}
+
+type Request struct {
+	ConversationId string    `json:"conversation_id,omitempty"`
+	BotId          string    `json:"bot_id"`
+	User           string    `json:"user"`
+	Query          string    `json:"query"`
+	ChatHistory    []Message `json:"chat_history,omitempty"`
+	Stream         bool      `json:"stream"`
+}
+
+type Response struct {
+	ConversationId string    `json:"conversation_id,omitempty"`
+	Messages       []Message `json:"messages,omitempty"`
+	Code           int       `json:"code,omitempty"`
+	Msg            string    `json:"msg,omitempty"`
+}
+
+type StreamResponse struct {
+	Event            string            `json:"event,omitempty"`
+	Message          *Message          `json:"message,omitempty"`
+	IsFinish         bool              `json:"is_finish,omitempty"`
+	Index            int               `json:"index,omitempty"`
+	ConversationId   string            `json:"conversation_id,omitempty"`
+	ErrorInformation *ErrorInformation `json:"error_information,omitempty"`
+}

relay/adaptor/deepl/adaptor.go

@@ -0,0 +1,73 @@
+package deepl
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/adaptor"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/model"
+	"io"
+	"net/http"
+)
+
+type Adaptor struct {
+	meta       *meta.Meta
+	promptText string
+}
+
+func (a *Adaptor) Init(meta *meta.Meta) {
+	a.meta = meta
+}
+
+func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
+	return fmt.Sprintf("%s/v2/translate", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *meta.Meta) error {
+	adaptor.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "DeepL-Auth-Key "+meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	convertedRequest, text := ConvertRequest(*request)
+	a.promptText = text
+	return convertedRequest, nil
+}
+
+func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return request, nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Reader) (*http.Response, error) {
+	return adaptor.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err = StreamHandler(c, resp, meta.ActualModelName)
+	} else {
+		err = Handler(c, resp, meta.ActualModelName)
+	}
+	promptTokens := len(a.promptText)
+	usage = &model.Usage{
+		PromptTokens: promptTokens,
+		TotalTokens:  promptTokens,
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "deepl"
+}

relay/adaptor/deepl/constants.go

@@ -0,0 +1,9 @@
+package deepl
+
+// https://developers.deepl.com/docs/api-reference/glossaries
+
+var ModelList = []string{
+	"deepl-zh",
+	"deepl-en",
+	"deepl-ja",
+}

relay/adaptor/deepl/helper.go

@@ -0,0 +1,11 @@
+package deepl
+
+import "strings"
+
+func parseLangFromModelName(modelName string) string {
+	parts := strings.Split(modelName, "-")
+	if len(parts) == 1 {
+		return "ZH"
+	}
+	return parts[1]
+}

relay/adaptor/deepl/main.go

@@ -0,0 +1,137 @@
+package deepl
+
+import (
+	"encoding/json"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/constant/finishreason"
+	"github.com/songquanpeng/one-api/relay/constant/role"
+	"github.com/songquanpeng/one-api/relay/model"
+	"io"
+	"net/http"
+)
+
+// https://developers.deepl.com/docs/getting-started/your-first-api-request
+
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) (*Request, string) {
+	var text string
+	if len(textRequest.Messages) != 0 {
+		text = textRequest.Messages[len(textRequest.Messages)-1].StringContent()
+	}
+	deeplRequest := Request{
+		TargetLang: parseLangFromModelName(textRequest.Model),
+		Text:       []string{text},
+	}
+	return &deeplRequest, text
+}
+
+func StreamResponseDeepL2OpenAI(deeplResponse *Response) *openai.ChatCompletionsStreamResponse {
+	var choice openai.ChatCompletionsStreamResponseChoice
+	if len(deeplResponse.Translations) != 0 {
+		choice.Delta.Content = deeplResponse.Translations[0].Text
+	}
+	choice.Delta.Role = role.Assistant
+	choice.FinishReason = &constant.StopFinishReason
+	openaiResponse := openai.ChatCompletionsStreamResponse{
+		Object:  constant.StreamObject,
+		Created: helper.GetTimestamp(),
+		Choices: []openai.ChatCompletionsStreamResponseChoice{choice},
+	}
+	return &openaiResponse
+}
+
+func ResponseDeepL2OpenAI(deeplResponse *Response) *openai.TextResponse {
+	var responseText string
+	if len(deeplResponse.Translations) != 0 {
+		responseText = deeplResponse.Translations[0].Text
+	}
+	choice := openai.TextResponseChoice{
+		Index: 0,
+		Message: model.Message{
+			Role:    role.Assistant,
+			Content: responseText,
+			Name:    nil,
+		},
+		FinishReason: finishreason.Stop,
+	}
+	fullTextResponse := openai.TextResponse{
+		Object:  constant.NonStreamObject,
+		Created: helper.GetTimestamp(),
+		Choices: []openai.TextResponseChoice{choice},
+	}
+	return &fullTextResponse
+}
+
+func StreamHandler(c *gin.Context, resp *http.Response, modelName string) *model.ErrorWithStatusCode {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError)
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError)
+	}
+	var deeplResponse Response
+	err = json.Unmarshal(responseBody, &deeplResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError)
+	}
+	fullTextResponse := StreamResponseDeepL2OpenAI(&deeplResponse)
+	fullTextResponse.Model = modelName
+	fullTextResponse.Id = helper.GetResponseID(c)
+	jsonData, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError)
+	}
+	common.SetEventStreamHeaders(c)
+	c.Stream(func(w io.Writer) bool {
+		if jsonData != nil {
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonData)})
+			jsonData = nil
+			return true
+		}
+		c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+		return false
+	})
+	_ = resp.Body.Close()
+	return nil
+}
+
+func Handler(c *gin.Context, resp *http.Response, modelName string) *model.ErrorWithStatusCode {
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError)
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError)
+	}
+	var deeplResponse Response
+	err = json.Unmarshal(responseBody, &deeplResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError)
+	}
+	if deeplResponse.Message != "" {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: deeplResponse.Message,
+				Code:    "deepl_error",
+			},
+			StatusCode: resp.StatusCode,
+		}
+	}
+	fullTextResponse := ResponseDeepL2OpenAI(&deeplResponse)
+	fullTextResponse.Model = modelName
+	fullTextResponse.Id = helper.GetResponseID(c)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError)
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil
+}

relay/adaptor/deepl/model.go

@@ -0,0 +1,16 @@
+package deepl
+
+type Request struct {
+	Text       []string `json:"text"`
+	TargetLang string   `json:"target_lang"`
+}
+
+type Translation struct {
+	DetectedSourceLanguage string `json:"detected_source_language,omitempty"`
+	Text                   string `json:"text,omitempty"`
+}
+
+type Response struct {
+	Translations []Translation `json:"translations,omitempty"`
+	Message      string        `json:"message,omitempty"`
+}

relay/adaptor/deepseek/constants.go

@@ -0,0 +1,6 @@
+package deepseek
+
+var ModelList = []string{
+	"deepseek-chat",
+	"deepseek-coder",
+}

relay/adaptor/doubao/constants.go

@@ -0,0 +1,13 @@
+package doubao
+
+// https://console.volcengine.com/ark/region:ark+cn-beijing/model
+
+var ModelList = []string{
+	"Doubao-pro-128k",
+	"Doubao-pro-32k",
+	"Doubao-pro-4k",
+	"Doubao-lite-128k",
+	"Doubao-lite-32k",
+	"Doubao-lite-4k",
+	"Doubao-embedding",
+}

relay/adaptor/doubao/main.go

@@ -0,0 +1,14 @@
+package doubao
+
+import (
+	"fmt"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/relaymode"
+)
+
+func GetRequestURL(meta *meta.Meta) (string, error) {
+	if meta.Mode == relaymode.ChatCompletions {
+		return fmt.Sprintf("%s/api/v3/chat/completions", meta.BaseURL), nil
+	}
+	return "", fmt.Errorf("unsupported relay mode %d for doubao", meta.Mode)
+}

relay/adaptor/gemini/adaptor.go

@@ -3,6 +3,9 @@ package gemini
 import (
 	"errors"
 	"fmt"
+	"io"
+	"net/http"
+
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
@@ -10,8 +13,7 @@ import (
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"github.com/songquanpeng/one-api/relay/model"
-	"io"
-	"net/http"
+	"github.com/songquanpeng/one-api/relay/relaymode"
 )
 
 type Adaptor struct {
@@ -22,10 +24,17 @@ func (a *Adaptor) Init(meta *meta.Meta) {
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
-	version := helper.AssignOrDefault(meta.APIVersion, config.GeminiVersion)
-	action := "generateContent"
+	version := helper.AssignOrDefault(meta.Config.APIVersion, config.GeminiVersion)
+	action := ""
+	switch meta.Mode {
+	case relaymode.Embeddings:
+		action = "batchEmbedContents"
+	default:
+		action = "generateContent"
+	}
+
 	if meta.IsStream {
-		action = "streamGenerateContent"
+		action = "streamGenerateContent?alt=sse"
 	}
 	return fmt.Sprintf("%s/%s/models/%s:%s", meta.BaseURL, version, meta.ActualModelName, action), nil
 }
@@ -40,7 +49,14 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	return ConvertRequest(*request), nil
+	switch relayMode {
+	case relaymode.Embeddings:
+		geminiEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return geminiEmbeddingRequest, nil
+	default:
+		geminiRequest := ConvertRequest(*request)
+		return geminiRequest, nil
+	}
 }
 
 func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
@@ -60,7 +76,12 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Met
 		err, responseText = StreamHandler(c, resp)
 		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
 	} else {
-		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+		switch meta.Mode {
+		case relaymode.Embeddings:
+			err, usage = EmbeddingHandler(c, resp)
+		default:
+			err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+		}
 	}
 	return
 }

relay/adaptor/gemini/constants.go

@@ -4,5 +4,5 @@ package gemini
 
 var ModelList = []string{
 	"gemini-pro", "gemini-1.0-pro-001", "gemini-1.5-pro",
-	"gemini-pro-vision", "gemini-1.0-pro-vision-001",
+	"gemini-pro-vision", "gemini-1.0-pro-vision-001", "embedding-001", "text-embedding-004",
 }

relay/adaptor/gemini/main.go

@@ -4,6 +4,10 @@ import (
 	"bufio"
 	"encoding/json"
 	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
@@ -13,9 +17,6 @@ import (
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/model"
-	"io"
-	"net/http"
-	"strings"
 
 	"github.com/gin-gonic/gin"
 )
@@ -54,7 +55,17 @@ func ConvertRequest(textRequest model.GeneralOpenAIRequest) *ChatRequest {
 			MaxOutputTokens: textRequest.MaxTokens,
 		},
 	}
-	if textRequest.Functions != nil {
+	if textRequest.Tools != nil {
+		functions := make([]model.Function, 0, len(textRequest.Tools))
+		for _, tool := range textRequest.Tools {
+			functions = append(functions, tool.Function)
+		}
+		geminiRequest.Tools = []ChatTools{
+			{
+				FunctionDeclarations: functions,
+			},
+		}
+	} else if textRequest.Functions != nil {
 		geminiRequest.Tools = []ChatTools{
 			{
 				FunctionDeclarations: textRequest.Functions,
@@ -123,6 +134,29 @@ func ConvertRequest(textRequest model.GeneralOpenAIRequest) *ChatRequest {
 	return &geminiRequest
 }
 
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *BatchEmbeddingRequest {
+	inputs := request.ParseInput()
+	requests := make([]EmbeddingRequest, len(inputs))
+	model := fmt.Sprintf("models/%s", request.Model)
+
+	for i, input := range inputs {
+		requests[i] = EmbeddingRequest{
+			Model: model,
+			Content: ChatContent{
+				Parts: []Part{
+					{
+						Text: input,
+					},
+				},
+			},
+		}
+	}
+
+	return &BatchEmbeddingRequest{
+		Requests: requests,
+	}
+}
+
 type ChatResponse struct {
 	Candidates     []ChatCandidate    `json:"candidates"`
 	PromptFeedback ChatPromptFeedback `json:"promptFeedback"`
@@ -154,6 +188,30 @@ type ChatPromptFeedback struct {
 	SafetyRatings []ChatSafetyRating `json:"safetyRatings"`
 }
 
+func getToolCalls(candidate *ChatCandidate) []model.Tool {
+	var toolCalls []model.Tool
+
+	item := candidate.Content.Parts[0]
+	if item.FunctionCall == nil {
+		return toolCalls
+	}
+	argsBytes, err := json.Marshal(item.FunctionCall.Arguments)
+	if err != nil {
+		logger.FatalLog("getToolCalls failed: " + err.Error())
+		return toolCalls
+	}
+	toolCall := model.Tool{
+		Id:   fmt.Sprintf("call_%s", random.GetUUID()),
+		Type: "function",
+		Function: model.Function{
+			Arguments: string(argsBytes),
+			Name:      item.FunctionCall.FunctionName,
+		},
+	}
+	toolCalls = append(toolCalls, toolCall)
+	return toolCalls
+}
+
 func responseGeminiChat2OpenAI(response *ChatResponse) *openai.TextResponse {
 	fullTextResponse := openai.TextResponse{
 		Id:      fmt.Sprintf("chatcmpl-%s", random.GetUUID()),
@@ -165,13 +223,19 @@ func responseGeminiChat2OpenAI(response *ChatResponse) *openai.TextResponse {
 		choice := openai.TextResponseChoice{
 			Index: i,
 			Message: model.Message{
-				Role:    "assistant",
-				Content: "",
+				Role: "assistant",
 			},
 			FinishReason: constant.StopFinishReason,
 		}
 		if len(candidate.Content.Parts) > 0 {
-			choice.Message.Content = candidate.Content.Parts[0].Text
+			if candidate.Content.Parts[0].FunctionCall != nil {
+				choice.Message.ToolCalls = getToolCalls(&candidate)
+			} else {
+				choice.Message.Content = candidate.Content.Parts[0].Text
+			}
+		} else {
+			choice.Message.Content = ""
+			choice.FinishReason = candidate.FinishReason
 		}
 		fullTextResponse.Choices = append(fullTextResponse.Choices, choice)
 	}
@@ -181,18 +245,35 @@ func responseGeminiChat2OpenAI(response *ChatResponse) *openai.TextResponse {
 func streamResponseGeminiChat2OpenAI(geminiResponse *ChatResponse) *openai.ChatCompletionsStreamResponse {
 	var choice openai.ChatCompletionsStreamResponseChoice
 	choice.Delta.Content = geminiResponse.GetResponseText()
-	choice.FinishReason = &constant.StopFinishReason
+	//choice.FinishReason = &constant.StopFinishReason
 	var response openai.ChatCompletionsStreamResponse
+	response.Id = fmt.Sprintf("chatcmpl-%s", random.GetUUID())
+	response.Created = helper.GetTimestamp()
 	response.Object = "chat.completion.chunk"
 	response.Model = "gemini"
 	response.Choices = []openai.ChatCompletionsStreamResponseChoice{choice}
 	return &response
 }
 
+func embeddingResponseGemini2OpenAI(response *EmbeddingResponse) *openai.EmbeddingResponse {
+	openAIEmbeddingResponse := openai.EmbeddingResponse{
+		Object: "list",
+		Data:   make([]openai.EmbeddingResponseItem, 0, len(response.Embeddings)),
+		Model:  "gemini-embedding",
+		Usage:  model.Usage{TotalTokens: 0},
+	}
+	for _, item := range response.Embeddings {
+		openAIEmbeddingResponse.Data = append(openAIEmbeddingResponse.Data, openai.EmbeddingResponseItem{
+			Object:    `embedding`,
+			Index:     0,
+			Embedding: item.Values,
+		})
+	}
+	return &openAIEmbeddingResponse
+}
+
 func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, string) {
 	responseText := ""
-	dataChan := make(chan string)
-	stopChan := make(chan bool)
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
@@ -206,14 +287,16 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 		}
 		return 0, nil, nil
 	})
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
 	go func() {
 		for scanner.Scan() {
 			data := scanner.Text()
 			data = strings.TrimSpace(data)
-			if !strings.HasPrefix(data, "\"text\": \"") {
+			if !strings.HasPrefix(data, "data: ") {
 				continue
 			}
-			data = strings.TrimPrefix(data, "\"text\": \"")
+			data = strings.TrimPrefix(data, "data: ")
 			data = strings.TrimSuffix(data, "\"")
 			dataChan <- data
 		}
@@ -223,23 +306,17 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:
-			// this is used to prevent annoying \ related format bug
-			data = fmt.Sprintf("{\"content\": \"%s\"}", data)
-			type dummyStruct struct {
-				Content string `json:"content"`
+			var geminiResponse ChatResponse
+			err := json.Unmarshal([]byte(data), &geminiResponse)
+			if err != nil {
+				logger.SysError("error unmarshalling stream response: " + err.Error())
+				return true
 			}
-			var dummy dummyStruct
-			err := json.Unmarshal([]byte(data), &dummy)
-			responseText += dummy.Content
-			var choice openai.ChatCompletionsStreamResponseChoice
-			choice.Delta.Content = dummy.Content
-			response := openai.ChatCompletionsStreamResponse{
-				Id:      fmt.Sprintf("chatcmpl-%s", random.GetUUID()),
-				Object:  "chat.completion.chunk",
-				Created: helper.GetTimestamp(),
-				Model:   "gemini-pro",
-				Choices: []openai.ChatCompletionsStreamResponseChoice{choice},
+			response := streamResponseGeminiChat2OpenAI(&geminiResponse)
+			if response == nil {
+				return true
 			}
+			responseText += response.Choices[0].Delta.StringContent()
 			jsonResponse, err := json.Marshal(response)
 			if err != nil {
 				logger.SysError("error marshalling stream response: " + err.Error())
@@ -302,3 +379,39 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 	_, err = c.Writer.Write(jsonResponse)
 	return nil, &usage
 }
+
+func EmbeddingHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var geminiEmbeddingResponse EmbeddingResponse
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = json.Unmarshal(responseBody, &geminiEmbeddingResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	if geminiEmbeddingResponse.Error != nil {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: geminiEmbeddingResponse.Error.Message,
+				Type:    "gemini_error",
+				Param:   "",
+				Code:    geminiEmbeddingResponse.Error.Code,
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+	fullTextResponse := embeddingResponseGemini2OpenAI(&geminiEmbeddingResponse)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &fullTextResponse.Usage
+}

relay/adaptor/gemini/model.go

@@ -7,14 +7,47 @@ type ChatRequest struct {
 	Tools            []ChatTools          `json:"tools,omitempty"`
 }
 
+type EmbeddingRequest struct {
+	Model                string      `json:"model"`
+	Content              ChatContent `json:"content"`
+	TaskType             string      `json:"taskType,omitempty"`
+	Title                string      `json:"title,omitempty"`
+	OutputDimensionality int         `json:"outputDimensionality,omitempty"`
+}
+
+type BatchEmbeddingRequest struct {
+	Requests []EmbeddingRequest `json:"requests"`
+}
+
+type EmbeddingData struct {
+	Values []float64 `json:"values"`
+}
+
+type EmbeddingResponse struct {
+	Embeddings []EmbeddingData `json:"embeddings"`
+	Error      *Error          `json:"error,omitempty"`
+}
+
+type Error struct {
+	Code    int    `json:"code,omitempty"`
+	Message string `json:"message,omitempty"`
+	Status  string `json:"status,omitempty"`
+}
+
 type InlineData struct {
 	MimeType string `json:"mimeType"`
 	Data     string `json:"data"`
 }
 
+type FunctionCall struct {
+	FunctionName string `json:"name"`
+	Arguments    any    `json:"args"`
+}
+
 type Part struct {
-	Text       string      `json:"text,omitempty"`
-	InlineData *InlineData `json:"inlineData,omitempty"`
+	Text         string        `json:"text,omitempty"`
+	InlineData   *InlineData   `json:"inlineData,omitempty"`
+	FunctionCall *FunctionCall `json:"functionCall,omitempty"`
 }
 
 type ChatContent struct {
@@ -28,7 +61,7 @@ type ChatSafetySettings struct {
 }
 
 type ChatTools struct {
-	FunctionDeclarations any `json:"functionDeclarations,omitempty"`
+	FunctionDeclarations any `json:"function_declarations,omitempty"`
 }
 
 type ChatGenerationConfig struct {

relay/adaptor/groq/constants.go

@@ -7,4 +7,6 @@ var ModelList = []string{
 	"llama2-7b-2048",
 	"llama2-70b-4096",
 	"mixtral-8x7b-32768",
+	"llama3-8b-8192",
+	"llama3-70b-8192",
 }

relay/adaptor/minimax/constants.go

@@ -1,7 +1,11 @@
 package minimax
 
+// https://www.minimaxi.com/document/guides/chat-model/V2?id=65e0736ab2845de20908e2dd
+
 var ModelList = []string{
-	"abab5.5s-chat",
-	"abab5.5-chat",
+	"abab6.5-chat",
+	"abab6.5s-chat",
 	"abab6-chat",
+	"abab5.5-chat",
+	"abab5.5s-chat",
 }

relay/adaptor/ollama/constants.go

@@ -1,5 +1,11 @@
 package ollama
 
 var ModelList = []string{
+	"codellama:7b-instruct",
+	"llama2:7b",
+	"llama2:latest",
+	"llama3:latest",
+	"phi3:latest",
 	"qwen:0.5b-chat",
+	"qwen:7b",
 }

relay/adaptor/ollama/main.go

@@ -13,6 +13,7 @@ import (
 
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/image"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/constant"
@@ -32,9 +33,22 @@ func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
 		Stream: request.Stream,
 	}
 	for _, message := range request.Messages {
+		openaiContent := message.ParseContent()
+		var imageUrls []string
+		var contentText string
+		for _, part := range openaiContent {
+			switch part.Type {
+			case model.ContentTypeText:
+				contentText = part.Text
+			case model.ContentTypeImageURL:
+				_, data, _ := image.GetImageFromUrl(part.ImageURL.Url)
+				imageUrls = append(imageUrls, data)
+			}
+		}
 		ollamaRequest.Messages = append(ollamaRequest.Messages, Message{
 			Role:    message.Role,
-			Content: message.StringContent(),
+			Content: contentText,
+			Images:  imageUrls,
 		})
 	}
 	return &ollamaRequest
@@ -53,6 +67,7 @@ func responseOllama2OpenAI(response *ChatResponse) *openai.TextResponse {
 	}
 	fullTextResponse := openai.TextResponse{
 		Id:      fmt.Sprintf("chatcmpl-%s", random.GetUUID()),
+		Model:   response.Model,
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
 		Choices: []openai.TextResponseChoice{choice},

relay/adaptor/openai/adaptor.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/relay/adaptor"
+	"github.com/songquanpeng/one-api/relay/adaptor/doubao"
 	"github.com/songquanpeng/one-api/relay/adaptor/minimax"
 	"github.com/songquanpeng/one-api/relay/channeltype"
 	"github.com/songquanpeng/one-api/relay/meta"
@@ -29,13 +30,13 @@ func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
 		if meta.Mode == relaymode.ImagesGenerations {
 			// https://learn.microsoft.com/en-us/azure/ai-services/openai/dall-e-quickstart?tabs=dalle3%2Ccommand-line&pivots=rest-api
 			// https://{resource_name}.openai.azure.com/openai/deployments/dall-e-3/images/generations?api-version=2024-03-01-preview
-			fullRequestURL := fmt.Sprintf("%s/openai/deployments/%s/images/generations?api-version=%s", meta.BaseURL, meta.ActualModelName, meta.APIVersion)
+			fullRequestURL := fmt.Sprintf("%s/openai/deployments/%s/images/generations?api-version=%s", meta.BaseURL, meta.ActualModelName, meta.Config.APIVersion)
 			return fullRequestURL, nil
 		}
 
 		// https://learn.microsoft.com/en-us/azure/cognitive-services/openai/chatgpt-quickstart?pivots=rest-api&tabs=command-line#rest-api
 		requestURL := strings.Split(meta.RequestURLPath, "?")[0]
-		requestURL = fmt.Sprintf("%s?api-version=%s", requestURL, meta.APIVersion)
+		requestURL = fmt.Sprintf("%s?api-version=%s", requestURL, meta.Config.APIVersion)
 		task := strings.TrimPrefix(requestURL, "/v1/")
 		model_ := meta.ActualModelName
 		model_ = strings.Replace(model_, ".", "", -1)
@@ -45,6 +46,8 @@ func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
 		return GetFullRequestURL(meta.BaseURL, requestURL, meta.ChannelType), nil
 	case channeltype.Minimax:
 		return minimax.GetRequestURL(meta)
+	case channeltype.Doubao:
+		return doubao.GetRequestURL(meta)
 	default:
 		return GetFullRequestURL(meta.BaseURL, meta.RequestURLPath, meta.ChannelType), nil
 	}
@@ -86,9 +89,13 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Met
 	if meta.IsStream {
 		var responseText string
 		err, responseText, usage = StreamHandler(c, resp, meta.Mode)
-		if usage == nil {
+		if usage == nil || usage.TotalTokens == 0 {
 			usage = ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
 		}
+		if usage.TotalTokens != 0 && usage.PromptTokens == 0 { // some channels don't return prompt tokens & completion tokens
+			usage.PromptTokens = meta.PromptTokens
+			usage.CompletionTokens = usage.TotalTokens - meta.PromptTokens
+		}
 	} else {
 		switch meta.Mode {
 		case relaymode.ImagesGenerations:

relay/adaptor/openai/compatible.go

@@ -3,12 +3,15 @@ package openai
 import (
 	"github.com/songquanpeng/one-api/relay/adaptor/ai360"
 	"github.com/songquanpeng/one-api/relay/adaptor/baichuan"
+	"github.com/songquanpeng/one-api/relay/adaptor/deepseek"
+	"github.com/songquanpeng/one-api/relay/adaptor/doubao"
 	"github.com/songquanpeng/one-api/relay/adaptor/groq"
 	"github.com/songquanpeng/one-api/relay/adaptor/lingyiwanwu"
 	"github.com/songquanpeng/one-api/relay/adaptor/minimax"
 	"github.com/songquanpeng/one-api/relay/adaptor/mistral"
 	"github.com/songquanpeng/one-api/relay/adaptor/moonshot"
 	"github.com/songquanpeng/one-api/relay/adaptor/stepfun"
+	"github.com/songquanpeng/one-api/relay/adaptor/togetherai"
 	"github.com/songquanpeng/one-api/relay/channeltype"
 )
 
@@ -18,10 +21,13 @@ var CompatibleChannels = []int{
 	channeltype.Moonshot,
 	channeltype.Baichuan,
 	channeltype.Minimax,
+	channeltype.Doubao,
 	channeltype.Mistral,
 	channeltype.Groq,
 	channeltype.LingYiWanWu,
 	channeltype.StepFun,
+	channeltype.DeepSeek,
+	channeltype.TogetherAI,
 }
 
 func GetCompatibleChannelMeta(channelType int) (string, []string) {
@@ -44,6 +50,12 @@ func GetCompatibleChannelMeta(channelType int) (string, []string) {
 		return "lingyiwanwu", lingyiwanwu.ModelList
 	case channeltype.StepFun:
 		return "stepfun", stepfun.ModelList
+	case channeltype.DeepSeek:
+		return "deepseek", deepseek.ModelList
+	case channeltype.TogetherAI:
+		return "together.ai", togetherai.ModelList
+	case channeltype.Doubao:
+		return "doubao", doubao.ModelList
 	default:
 		return "openai", ModelList
 	}

relay/adaptor/openai/constants.go

@@ -7,6 +7,7 @@ var ModelList = []string{
 	"gpt-4", "gpt-4-0314", "gpt-4-0613", "gpt-4-1106-preview", "gpt-4-0125-preview",
 	"gpt-4-32k", "gpt-4-32k-0314", "gpt-4-32k-0613",
 	"gpt-4-turbo-preview", "gpt-4-turbo", "gpt-4-turbo-2024-04-09",
+	"gpt-4o", "gpt-4o-2024-05-13",
 	"gpt-4-vision-preview",
 	"text-embedding-ada-002", "text-embedding-3-small", "text-embedding-3-large",
 	"text-curie-001", "text-babbage-001", "text-ada-001", "text-davinci-002", "text-davinci-003",

relay/adaptor/openai/main.go

@@ -15,6 +15,12 @@ import (
 	"strings"
 )
 
+const (
+	dataPrefix       = "data: "
+	done             = "[DONE]"
+	dataPrefixLength = len(dataPrefix)
+)
+
 func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.ErrorWithStatusCode, string, *model.Usage) {
 	responseText := ""
 	scanner := bufio.NewScanner(resp.Body)
@@ -36,39 +42,46 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.E
 	go func() {
 		for scanner.Scan() {
 			data := scanner.Text()
-			if len(data) < 6 { // ignore blank line or wrong format
+			if len(data) < dataPrefixLength { // ignore blank line or wrong format
 				continue
 			}
-			if data[:6] != "data: " && data[:6] != "[DONE]" {
+			if data[:dataPrefixLength] != dataPrefix && data[:dataPrefixLength] != done {
 				continue
 			}
-			dataChan <- data
-			data = data[6:]
-			if !strings.HasPrefix(data, "[DONE]") {
-				switch relayMode {
-				case relaymode.ChatCompletions:
-					var streamResponse ChatCompletionsStreamResponse
-					err := json.Unmarshal([]byte(data), &streamResponse)
-					if err != nil {
-						logger.SysError("error unmarshalling stream response: " + err.Error())
-						continue // just ignore the error
-					}
-					for _, choice := range streamResponse.Choices {
-						responseText += conv.AsString(choice.Delta.Content)
-					}
-					if streamResponse.Usage != nil {
-						usage = streamResponse.Usage
-					}
-				case relaymode.Completions:
-					var streamResponse CompletionsStreamResponse
-					err := json.Unmarshal([]byte(data), &streamResponse)
-					if err != nil {
-						logger.SysError("error unmarshalling stream response: " + err.Error())
-						continue
-					}
-					for _, choice := range streamResponse.Choices {
-						responseText += choice.Text
-					}
+			if strings.HasPrefix(data[dataPrefixLength:], done) {
+				dataChan <- data
+				continue
+			}
+			switch relayMode {
+			case relaymode.ChatCompletions:
+				var streamResponse ChatCompletionsStreamResponse
+				err := json.Unmarshal([]byte(data[dataPrefixLength:]), &streamResponse)
+				if err != nil {
+					logger.SysError("error unmarshalling stream response: " + err.Error())
+					dataChan <- data // if error happened, pass the data to client
+					continue         // just ignore the error
+				}
+				if len(streamResponse.Choices) == 0 {
+					// but for empty choice, we should not pass it to client, this is for azure
+					continue // just ignore empty choice
+				}
+				dataChan <- data
+				for _, choice := range streamResponse.Choices {
+					responseText += conv.AsString(choice.Delta.Content)
+				}
+				if streamResponse.Usage != nil {
+					usage = streamResponse.Usage
+				}
+			case relaymode.Completions:
+				dataChan <- data
+				var streamResponse CompletionsStreamResponse
+				err := json.Unmarshal([]byte(data[dataPrefixLength:]), &streamResponse)
+				if err != nil {
+					logger.SysError("error unmarshalling stream response: " + err.Error())
+					continue
+				}
+				for _, choice := range streamResponse.Choices {
+					responseText += choice.Text
 				}
 			}
 		}

relay/adaptor/openai/model.go

@@ -134,7 +134,7 @@ type ChatCompletionsStreamResponse struct {
 	Created int64                                 `json:"created"`
 	Model   string                                `json:"model"`
 	Choices []ChatCompletionsStreamResponseChoice `json:"choices"`
-	Usage   *model.Usage                          `json:"usage"`
+	Usage   *model.Usage                          `json:"usage,omitempty"`
 }
 
 type CompletionsStreamResponse struct {

relay/adaptor/openai/token.go

@@ -24,6 +24,10 @@ func InitTokenEncoders() {
 		logger.FatalLog(fmt.Sprintf("failed to get gpt-3.5-turbo token encoder: %s", err.Error()))
 	}
 	defaultTokenEncoder = gpt35TokenEncoder
+	gpt4oTokenEncoder, err := tiktoken.EncodingForModel("gpt-4o")
+	if err != nil {
+		logger.FatalLog(fmt.Sprintf("failed to get gpt-4o token encoder: %s", err.Error()))
+	}
 	gpt4TokenEncoder, err := tiktoken.EncodingForModel("gpt-4")
 	if err != nil {
 		logger.FatalLog(fmt.Sprintf("failed to get gpt-4 token encoder: %s", err.Error()))
@@ -31,6 +35,8 @@ func InitTokenEncoders() {
 	for model := range billingratio.ModelRatio {
 		if strings.HasPrefix(model, "gpt-3.5") {
 			tokenEncoderMap[model] = gpt35TokenEncoder
+		} else if strings.HasPrefix(model, "gpt-4o") {
+			tokenEncoderMap[model] = gpt4oTokenEncoder
 		} else if strings.HasPrefix(model, "gpt-4") {
 			tokenEncoderMap[model] = gpt4TokenEncoder
 		} else {
@@ -206,3 +212,7 @@ func CountTokenText(text string, model string) int {
 	tokenEncoder := getTokenEncoder(model)
 	return getTokenNum(tokenEncoder, text)
 }
+
+func CountToken(text string) int {
+	return CountTokenInput(text, "gpt-3.5-turbo")
+}

relay/adaptor/tencent/adaptor.go

@@ -2,35 +2,43 @@ package tencent
 
 import (
 	"errors"
-	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/relay/adaptor"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
+	"strconv"
 	"strings"
 )
 
 // https://cloud.tencent.com/document/api/1729/101837
 
 type Adaptor struct {
-	Sign string
+	Sign      string
+	Action    string
+	Version   string
+	Timestamp int64
 }
 
 func (a *Adaptor) Init(meta *meta.Meta) {
-
+	a.Action = "ChatCompletions"
+	a.Version = "2023-09-01"
+	a.Timestamp = helper.GetTimestamp()
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
-	return fmt.Sprintf("%s/hyllm/v1/chat/completions", meta.BaseURL), nil
+	return meta.BaseURL + "/", nil
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *meta.Meta) error {
 	adaptor.SetupCommonRequestHeader(c, req, meta)
 	req.Header.Set("Authorization", a.Sign)
-	req.Header.Set("X-TC-Action", meta.ActualModelName)
+	req.Header.Set("X-TC-Action", a.Action)
+	req.Header.Set("X-TC-Version", a.Version)
+	req.Header.Set("X-TC-Timestamp", strconv.FormatInt(a.Timestamp, 10))
 	return nil
 }
 
@@ -40,15 +48,13 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	}
 	apiKey := c.Request.Header.Get("Authorization")
 	apiKey = strings.TrimPrefix(apiKey, "Bearer ")
-	appId, secretId, secretKey, err := ParseConfig(apiKey)
+	_, secretId, secretKey, err := ParseConfig(apiKey)
 	if err != nil {
 		return nil, err
 	}
 	tencentRequest := ConvertRequest(*request)
-	tencentRequest.AppId = appId
-	tencentRequest.SecretId = secretId
 	// we have to calculate the sign here
-	a.Sign = GetSign(*tencentRequest, secretKey)
+	a.Sign = GetSign(*tencentRequest, a, secretId, secretKey)
 	return tencentRequest, nil
 }
 

relay/adaptor/tencent/constants.go

@@ -1,7 +1,8 @@
 package tencent
 
 var ModelList = []string{
-	"ChatPro",
-	"ChatStd",
-	"hunyuan",
+	"hunyuan-lite",
+	"hunyuan-standard",
+	"hunyuan-standard-256K",
+	"hunyuan-pro",
 }

relay/adaptor/tencent/main.go

@@ -3,8 +3,8 @@ package tencent
 import (
 	"bufio"
 	"crypto/hmac"
-	"crypto/sha1"
-	"encoding/base64"
+	"crypto/sha256"
+	"encoding/hex"
 	"encoding/json"
 	"errors"
 	"fmt"
@@ -19,34 +19,26 @@ import (
 	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"sort"
 	"strconv"
 	"strings"
+	"time"
 )
 
-// https://cloud.tencent.com/document/product/1729/97732
-
 func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
-	messages := make([]Message, 0, len(request.Messages))
+	messages := make([]*Message, 0, len(request.Messages))
 	for i := 0; i < len(request.Messages); i++ {
 		message := request.Messages[i]
-		messages = append(messages, Message{
+		messages = append(messages, &Message{
 			Content: message.StringContent(),
 			Role:    message.Role,
 		})
 	}
-	stream := 0
-	if request.Stream {
-		stream = 1
-	}
 	return &ChatRequest{
-		Timestamp:   helper.GetTimestamp(),
-		Expired:     helper.GetTimestamp() + 24*60*60,
-		QueryID:     random.GetUUID(),
-		Temperature: request.Temperature,
-		TopP:        request.TopP,
-		Stream:      stream,
+		Model:       &request.Model,
+		Stream:      &request.Stream,
 		Messages:    messages,
+		TopP:        &request.TopP,
+		Temperature: &request.Temperature,
 	}
 }
 
@@ -54,7 +46,11 @@ func responseTencent2OpenAI(response *ChatResponse) *openai.TextResponse {
 	fullTextResponse := openai.TextResponse{
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
-		Usage:   response.Usage,
+		Usage: model.Usage{
+			PromptTokens:     response.Usage.PromptTokens,
+			CompletionTokens: response.Usage.CompletionTokens,
+			TotalTokens:      response.Usage.TotalTokens,
+		},
 	}
 	if len(response.Choices) > 0 {
 		choice := openai.TextResponseChoice{
@@ -154,6 +150,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 
 func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var TencentResponse ChatResponse
+	var responseP ChatResponseP
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
@@ -162,10 +159,11 @@ func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *
 	if err != nil {
 		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
-	err = json.Unmarshal(responseBody, &TencentResponse)
+	err = json.Unmarshal(responseBody, &responseP)
 	if err != nil {
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
+	TencentResponse = responseP.Response
 	if TencentResponse.Error.Code != 0 {
 		return &model.ErrorWithStatusCode{
 			Error: model.Error{
@@ -202,29 +200,62 @@ func ParseConfig(config string) (appId int64, secretId string, secretKey string,
 	return
 }
 
-func GetSign(req ChatRequest, secretKey string) string {
-	params := make([]string, 0)
-	params = append(params, "app_id="+strconv.FormatInt(req.AppId, 10))
-	params = append(params, "secret_id="+req.SecretId)
-	params = append(params, "timestamp="+strconv.FormatInt(req.Timestamp, 10))
-	params = append(params, "query_id="+req.QueryID)
-	params = append(params, "temperature="+strconv.FormatFloat(req.Temperature, 'f', -1, 64))
-	params = append(params, "top_p="+strconv.FormatFloat(req.TopP, 'f', -1, 64))
-	params = append(params, "stream="+strconv.Itoa(req.Stream))
-	params = append(params, "expired="+strconv.FormatInt(req.Expired, 10))
-
-	var messageStr string
-	for _, msg := range req.Messages {
-		messageStr += fmt.Sprintf(`{"role":"%s","content":"%s"},`, msg.Role, msg.Content)
-	}
-	messageStr = strings.TrimSuffix(messageStr, ",")
-	params = append(params, "messages=["+messageStr+"]")
-
-	sort.Strings(params)
-	url := "hunyuan.cloud.tencent.com/hyllm/v1/chat/completions?" + strings.Join(params, "&")
-	mac := hmac.New(sha1.New, []byte(secretKey))
-	signURL := url
-	mac.Write([]byte(signURL))
-	sign := mac.Sum([]byte(nil))
-	return base64.StdEncoding.EncodeToString(sign)
+func sha256hex(s string) string {
+	b := sha256.Sum256([]byte(s))
+	return hex.EncodeToString(b[:])
+}
+
+func hmacSha256(s, key string) string {
+	hashed := hmac.New(sha256.New, []byte(key))
+	hashed.Write([]byte(s))
+	return string(hashed.Sum(nil))
+}
+
+func GetSign(req ChatRequest, adaptor *Adaptor, secId, secKey string) string {
+	// build canonical request string
+	host := "hunyuan.tencentcloudapi.com"
+	httpRequestMethod := "POST"
+	canonicalURI := "/"
+	canonicalQueryString := ""
+	canonicalHeaders := fmt.Sprintf("content-type:%s\nhost:%s\nx-tc-action:%s\n",
+		"application/json", host, strings.ToLower(adaptor.Action))
+	signedHeaders := "content-type;host;x-tc-action"
+	payload, _ := json.Marshal(req)
+	hashedRequestPayload := sha256hex(string(payload))
+	canonicalRequest := fmt.Sprintf("%s\n%s\n%s\n%s\n%s\n%s",
+		httpRequestMethod,
+		canonicalURI,
+		canonicalQueryString,
+		canonicalHeaders,
+		signedHeaders,
+		hashedRequestPayload)
+	// build string to sign
+	algorithm := "TC3-HMAC-SHA256"
+	requestTimestamp := strconv.FormatInt(adaptor.Timestamp, 10)
+	timestamp, _ := strconv.ParseInt(requestTimestamp, 10, 64)
+	t := time.Unix(timestamp, 0).UTC()
+	// must be the format 2006-01-02, ref to package time for more info
+	date := t.Format("2006-01-02")
+	credentialScope := fmt.Sprintf("%s/%s/tc3_request", date, "hunyuan")
+	hashedCanonicalRequest := sha256hex(canonicalRequest)
+	string2sign := fmt.Sprintf("%s\n%s\n%s\n%s",
+		algorithm,
+		requestTimestamp,
+		credentialScope,
+		hashedCanonicalRequest)
+
+	// sign string
+	secretDate := hmacSha256(date, "TC3"+secKey)
+	secretService := hmacSha256("hunyuan", secretDate)
+	secretKey := hmacSha256("tc3_request", secretService)
+	signature := hex.EncodeToString([]byte(hmacSha256(string2sign, secretKey)))
+
+	// build authorization
+	authorization := fmt.Sprintf("%s Credential=%s/%s, SignedHeaders=%s, Signature=%s",
+		algorithm,
+		secId,
+		credentialScope,
+		signedHeaders,
+		signature)
+	return authorization
 }

relay/adaptor/tencent/model.go

@@ -1,63 +1,75 @@
 package tencent
 
-import (
-	"github.com/songquanpeng/one-api/relay/model"
-)
-
 type Message struct {
-	Role    string `json:"role"`
-	Content string `json:"content"`
+	Role    string `json:"Role"`
+	Content string `json:"Content"`
 }
 
 type ChatRequest struct {
-	AppId    int64  `json:"app_id"`    // 腾讯云账号的 APPID
-	SecretId string `json:"secret_id"` // 官网 SecretId
-	// Timestamp当前 UNIX 时间戳，单位为秒，可记录发起 API 请求的时间。
-	// 例如1529223702，如果与当前时间相差过大，会引起签名过期错误
-	Timestamp int64 `json:"timestamp"`
-	// Expired 签名的有效期，是一个符合 UNIX Epoch 时间戳规范的数值，
-	// 单位为秒；Expired 必须大于 Timestamp 且 Expired-Timestamp 小于90天
-	Expired int64  `json:"expired"`
-	QueryID string `json:"query_id"` //请求 Id，用于问题排查
-	// Temperature 较高的数值会使输出更加随机，而较低的数值会使其更加集中和确定
-	// 默认 1.0，取值区间为[0.0,2.0]，非必要不建议使用,不合理的取值会影响效果
-	// 建议该参数和 top_p 只设置1个，不要同时更改 top_p
-	Temperature float64 `json:"temperature"`
-	// TopP 影响输出文本的多样性，取值越大，生成文本的多样性越强
-	// 默认1.0，取值区间为[0.0, 1.0]，非必要不建议使用, 不合理的取值会影响效果
-	// 建议该参数和 temperature 只设置1个，不要同时更改
-	TopP float64 `json:"top_p"`
-	// Stream 0：同步，1：流式 （默认，协议：SSE)
-	// 同步请求超时：60s，如果内容较长建议使用流式
-	Stream int `json:"stream"`
-	// Messages 会话内容, 长度最多为40, 按对话时间从旧到新在数组中排列
-	// 输入 content 总数最大支持 3000 token。
-	Messages []Message `json:"messages"`
+	// 模型名称，可选值包括 hunyuan-lite、hunyuan-standard、hunyuan-standard-256K、hunyuan-pro。
+	// 各模型介绍请阅读 [产品概述](https://cloud.tencent.com/document/product/1729/104753) 中的说明。
+	//
+	// 注意：
+	// 不同的模型计费不同，请根据 [购买指南](https://cloud.tencent.com/document/product/1729/97731) 按需调用。
+	Model *string `json:"Model"`
+	// 聊天上下文信息。
+	// 说明：
+	// 1. 长度最多为 40，按对话时间从旧到新在数组中排列。
+	// 2. Message.Role 可选值：system、user、assistant。
+	// 其中，system 角色可选，如存在则必须位于列表的最开始。user 和 assistant 需交替出现（一问一答），以 user 提问开始和结束，且 Content 不能为空。Role 的顺序示例：[system（可选） user assistant user assistant user ...]。
+	// 3. Messages 中 Content 总长度不能超过模型输入长度上限（可参考 [产品概述](https://cloud.tencent.com/document/product/1729/104753) 文档），超过则会截断最前面的内容，只保留尾部内容。
+	Messages []*Message `json:"Messages"`
+	// 流式调用开关。
+	// 说明：
+	// 1. 未传值时默认为非流式调用（false）。
+	// 2. 流式调用时以 SSE 协议增量返回结果（返回值取 Choices[n].Delta 中的值，需要拼接增量数据才能获得完整结果）。
+	// 3. 非流式调用时：
+	// 调用方式与普通 HTTP 请求无异。
+	// 接口响应耗时较长，**如需更低时延建议设置为 true**。
+	// 只返回一次最终结果（返回值取 Choices[n].Message 中的值）。
+	//
+	// 注意：
+	// 通过 SDK 调用时，流式和非流式调用需用**不同的方式**获取返回值，具体参考 SDK 中的注释或示例（在各语言 SDK 代码仓库的 examples/hunyuan/v20230901/ 目录中）。
+	Stream *bool `json:"Stream"`
+	// 说明：
+	// 1. 影响输出文本的多样性，取值越大，生成文本的多样性越强。
+	// 2. 取值区间为 [0.0, 1.0]，未传值时使用各模型推荐值。
+	// 3. 非必要不建议使用，不合理的取值会影响效果。
+	TopP *float64 `json:"TopP"`
+	// 说明：
+	// 1. 较高的数值会使输出更加随机，而较低的数值会使其更加集中和确定。
+	// 2. 取值区间为 [0.0, 2.0]，未传值时使用各模型推荐值。
+	// 3. 非必要不建议使用，不合理的取值会影响效果。
+	Temperature *float64 `json:"Temperature"`
 }
 
 type Error struct {
-	Code    int    `json:"code"`
-	Message string `json:"message"`
+	Code    int    `json:"Code"`
+	Message string `json:"Message"`
 }
 
 type Usage struct {
-	InputTokens  int `json:"input_tokens"`
-	OutputTokens int `json:"output_tokens"`
-	TotalTokens  int `json:"total_tokens"`
+	PromptTokens     int `json:"PromptTokens"`
+	CompletionTokens int `json:"CompletionTokens"`
+	TotalTokens      int `json:"TotalTokens"`
 }
 
 type ResponseChoices struct {
-	FinishReason string  `json:"finish_reason,omitempty"` // 流式结束标志位，为 stop 则表示尾包
-	Messages     Message `json:"messages,omitempty"`      // 内容，同步模式返回内容，流模式为 null 输出 content 内容总数最多支持 1024token。
-	Delta        Message `json:"delta,omitempty"`         // 内容，流模式返回内容，同步模式为 null 输出 content 内容总数最多支持 1024token。
+	FinishReason string  `json:"FinishReason,omitempty"` // 流式结束标志位，为 stop 则表示尾包
+	Messages     Message `json:"Message,omitempty"`      // 内容，同步模式返回内容，流模式为 null 输出 content 内容总数最多支持 1024token。
+	Delta        Message `json:"Delta,omitempty"`        // 内容，流模式返回内容，同步模式为 null 输出 content 内容总数最多支持 1024token。
 }
 
 type ChatResponse struct {
-	Choices []ResponseChoices `json:"choices,omitempty"` // 结果
-	Created string            `json:"created,omitempty"` // unix 时间戳的字符串
-	Id      string            `json:"id,omitempty"`      // 会话 id
-	Usage   model.Usage       `json:"usage,omitempty"`   // token 数量
-	Error   Error             `json:"error,omitempty"`   // 错误信息 注意：此字段可能返回 null，表示取不到有效值
-	Note    string            `json:"note,omitempty"`    // 注释
-	ReqID   string            `json:"req_id,omitempty"`  // 唯一请求 Id，每次请求都会返回。用于反馈接口入参
+	Choices []ResponseChoices `json:"Choices,omitempty"` // 结果
+	Created int64             `json:"Created,omitempty"` // unix 时间戳的字符串
+	Id      string            `json:"Id,omitempty"`      // 会话 id
+	Usage   Usage             `json:"Usage,omitempty"`   // token 数量
+	Error   Error             `json:"Error,omitempty"`   // 错误信息 注意：此字段可能返回 null，表示取不到有效值
+	Note    string            `json:"Note,omitempty"`    // 注释
+	ReqID   string            `json:"Req_id,omitempty"`  // 唯一请求 Id，每次请求都会返回。用于反馈接口入参
+}
+
+type ChatResponseP struct {
+	Response ChatResponse `json:"Response,omitempty"`
 }

relay/adaptor/togetherai/constants.go

@@ -0,0 +1,10 @@
+package togetherai
+
+// https://docs.together.ai/docs/inference-models
+
+var ModelList = []string{
+	"meta-llama/Llama-3-70b-chat-hf",
+	"deepseek-ai/deepseek-coder-33b-instruct",
+	"mistralai/Mixtral-8x22B-Instruct-v0.1",
+	"Qwen/Qwen1.5-72B-Chat",
+}

relay/adaptor/xunfei/adaptor.go

@@ -14,10 +14,11 @@ import (
 
 type Adaptor struct {
 	request *model.GeneralOpenAIRequest
+	meta    *meta.Meta
 }
 
 func (a *Adaptor) Init(meta *meta.Meta) {
-
+	a.meta = meta
 }
 
 func (a *Adaptor) GetRequestURL(meta *meta.Meta) (string, error) {
@@ -60,10 +61,18 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Met
 	if a.request == nil {
 		return nil, openai.ErrorWrapper(errors.New("request is nil"), "request_is_nil", http.StatusBadRequest)
 	}
+	version := parseAPIVersionByModelName(meta.ActualModelName)
+	if version == "" {
+		version = a.meta.Config.APIVersion
+	}
+	if version == "" {
+		version = "v1.1"
+	}
+	a.meta.Config.APIVersion = version
 	if meta.IsStream {
-		err, usage = StreamHandler(c, *a.request, splits[0], splits[1], splits[2])
+		err, usage = StreamHandler(c, meta, *a.request, splits[0], splits[1], splits[2])
 	} else {
-		err, usage = Handler(c, *a.request, splits[0], splits[1], splits[2])
+		err, usage = Handler(c, meta, *a.request, splits[0], splits[1], splits[2])
 	}
 	return
 }

relay/adaptor/xunfei/main.go

@@ -5,22 +5,24 @@ import (
 	"crypto/sha256"
 	"encoding/base64"
 	"encoding/json"
+	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
-	"github.com/gorilla/websocket"
-	"github.com/songquanpeng/one-api/common"
-	"github.com/songquanpeng/one-api/common/config"
-	"github.com/songquanpeng/one-api/common/helper"
-	"github.com/songquanpeng/one-api/common/logger"
-	"github.com/songquanpeng/one-api/common/random"
-	"github.com/songquanpeng/one-api/relay/adaptor/openai"
-	"github.com/songquanpeng/one-api/relay/constant"
-	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
 	"net/url"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
+	"github.com/gorilla/websocket"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/random"
+	"github.com/songquanpeng/one-api/relay/adaptor/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/meta"
+	"github.com/songquanpeng/one-api/relay/model"
 )
 
 // https://console.xfyun.cn/services/cbm
@@ -28,11 +30,7 @@ import (
 
 func requestOpenAI2Xunfei(request model.GeneralOpenAIRequest, xunfeiAppId string, domain string) *ChatRequest {
 	messages := make([]Message, 0, len(request.Messages))
-	var lastToolCalls []model.Tool
 	for _, message := range request.Messages {
-		if message.ToolCalls != nil {
-			lastToolCalls = message.ToolCalls
-		}
 		messages = append(messages, Message{
 			Role:    message.Role,
 			Content: message.StringContent(),
@@ -45,9 +43,10 @@ func requestOpenAI2Xunfei(request model.GeneralOpenAIRequest, xunfeiAppId string
 	xunfeiRequest.Parameter.Chat.TopK = request.N
 	xunfeiRequest.Parameter.Chat.MaxTokens = request.MaxTokens
 	xunfeiRequest.Payload.Message.Text = messages
-	if len(lastToolCalls) != 0 {
-		for _, toolCall := range lastToolCalls {
-			xunfeiRequest.Payload.Functions.Text = append(xunfeiRequest.Payload.Functions.Text, toolCall.Function)
+
+	if strings.HasPrefix(domain, "generalv3") {
+		xunfeiRequest.Payload.Functions = &Functions{
+			Text: request.Tools,
 		}
 	}
 
@@ -149,8 +148,8 @@ func buildXunfeiAuthUrl(hostUrl string, apiKey, apiSecret string) string {
 	return callUrl
 }
 
-func StreamHandler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId string, apiSecret string, apiKey string) (*model.ErrorWithStatusCode, *model.Usage) {
-	domain, authUrl := getXunfeiAuthUrl(c, apiKey, apiSecret, textRequest.Model)
+func StreamHandler(c *gin.Context, meta *meta.Meta, textRequest model.GeneralOpenAIRequest, appId string, apiSecret string, apiKey string) (*model.ErrorWithStatusCode, *model.Usage) {
+	domain, authUrl := getXunfeiAuthUrl(meta.Config.APIVersion, apiKey, apiSecret)
 	dataChan, stopChan, err := xunfeiMakeRequest(textRequest, domain, authUrl, appId)
 	if err != nil {
 		return openai.ErrorWrapper(err, "xunfei_request_failed", http.StatusInternalServerError), nil
@@ -179,8 +178,8 @@ func StreamHandler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId
 	return nil, &usage
 }
 
-func Handler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId string, apiSecret string, apiKey string) (*model.ErrorWithStatusCode, *model.Usage) {
-	domain, authUrl := getXunfeiAuthUrl(c, apiKey, apiSecret, textRequest.Model)
+func Handler(c *gin.Context, meta *meta.Meta, textRequest model.GeneralOpenAIRequest, appId string, apiSecret string, apiKey string) (*model.ErrorWithStatusCode, *model.Usage) {
+	domain, authUrl := getXunfeiAuthUrl(meta.Config.APIVersion, apiKey, apiSecret)
 	dataChan, stopChan, err := xunfeiMakeRequest(textRequest, domain, authUrl, appId)
 	if err != nil {
 		return openai.ErrorWrapper(err, "xunfei_request_failed", http.StatusInternalServerError), nil
@@ -203,7 +202,7 @@ func Handler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId strin
 		}
 	}
 	if len(xunfeiResponse.Payload.Choices.Text) == 0 {
-		return openai.ErrorWrapper(err, "xunfei_empty_response_detected", http.StatusInternalServerError), nil
+		return openai.ErrorWrapper(errors.New("xunfei empty response detected"), "xunfei_empty_response_detected", http.StatusInternalServerError), nil
 	}
 	xunfeiResponse.Payload.Choices.Text[0].Content = content
 
@@ -268,25 +267,12 @@ func xunfeiMakeRequest(textRequest model.GeneralOpenAIRequest, domain, authUrl,
 	return dataChan, stopChan, nil
 }
 
-func getAPIVersion(c *gin.Context, modelName string) string {
-	query := c.Request.URL.Query()
-	apiVersion := query.Get("api-version")
-	if apiVersion != "" {
-		return apiVersion
-	}
+func parseAPIVersionByModelName(modelName string) string {
 	parts := strings.Split(modelName, "-")
 	if len(parts) == 2 {
-		apiVersion = parts[1]
-		return apiVersion
-
+		return parts[1]
 	}
-	apiVersion = c.GetString(config.KeyAPIVersion)
-	if apiVersion != "" {
-		return apiVersion
-	}
-	apiVersion = "v1.1"
-	logger.SysLog("api_version not found, using default: " + apiVersion)
-	return apiVersion
+	return ""
 }
 
 // https://www.xfyun.cn/doc/spark/Web.html#_1-%E6%8E%A5%E5%8F%A3%E8%AF%B4%E6%98%8E
@@ -304,8 +290,7 @@ func apiVersion2domain(apiVersion string) string {
 	return "general" + apiVersion
 }
 
-func getXunfeiAuthUrl(c *gin.Context, apiKey string, apiSecret string, modelName string) (string, string) {
-	apiVersion := getAPIVersion(c, modelName)
+func getXunfeiAuthUrl(apiVersion string, apiKey string, apiSecret string) (string, string) {
 	domain := apiVersion2domain(apiVersion)
 	authUrl := buildXunfeiAuthUrl(fmt.Sprintf("wss://spark-api.xf-yun.com/%s/chat", apiVersion), apiKey, apiSecret)
 	return domain, authUrl

relay/adaptor/xunfei/model.go

@@ -9,6 +9,10 @@ type Message struct {
 	Content string `json:"content"`
 }
 
+type Functions struct {
+	Text []model.Tool `json:"text,omitempty"`
+}
+
 type ChatRequest struct {
 	Header struct {
 		AppId string `json:"app_id"`
@@ -26,9 +30,7 @@ type ChatRequest struct {
 		Message struct {
 			Text []Message `json:"text"`
 		} `json:"message"`
-		Functions struct {
-			Text []model.Function `json:"text,omitempty"`
-		} `json:"functions,omitempty"`
+		Functions *Functions `json:"functions,omitempty"`
 	} `json:"payload"`
 }
 

relay/adaptor/zhipu/adaptor.go

@@ -62,8 +62,8 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	}
 	switch relayMode {
 	case relaymode.Embeddings:
-		baiduEmbeddingRequest := ConvertEmbeddingRequest(*request)
-		return baiduEmbeddingRequest, nil
+		baiduEmbeddingRequest, err := ConvertEmbeddingRequest(*request)
+		return baiduEmbeddingRequest, err
 	default:
 		// TopP (0.0, 1.0)
 		request.TopP = math.Min(0.99, request.TopP)
@@ -129,11 +129,15 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Met
 	return
 }
 
-func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingRequest {
-	return &EmbeddingRequest{
-		Model: "embedding-2",
-		Input: request.Input.(string),
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) (*EmbeddingRequest, error) {
+	inputs := request.ParseInput()
+	if len(inputs) != 1 {
+		return nil, errors.New("invalid input length, zhipu only support one input")
 	}
+	return &EmbeddingRequest{
+		Model: request.Model,
+		Input: inputs[0],
+	}, nil
 }
 
 func (a *Adaptor) GetModelList() []string {

relay/apitype/define.go

@@ -13,6 +13,10 @@ const (
 	Gemini
 	Ollama
 	AwsClaude
+	Coze
+	Cohere
+	Cloudflare
+	DeepL
 
 	Dummy // this one is only for count, do not add any channel after this
 )

relay/billing/ratio/image.go

@@ -49,3 +49,8 @@ var ImagePromptLengthLimitations = map[string]int{
 	"wanx-v1":                   4000,
 	"cogview-3":                 833,
 }
+
+var ImageOriginModelName = map[string]string{
+	"ali-stable-diffusion-xl":   "stable-diffusion-xl",
+	"ali-stable-diffusion-v1.5": "stable-diffusion-v1.5",
+}

relay/billing/ratio/model.go

@@ -2,8 +2,9 @@ package ratio
 
 import (
 	"encoding/json"
-	"github.com/songquanpeng/one-api/common/logger"
 	"strings"
+
+	"github.com/songquanpeng/one-api/common/logger"
 )
 
 const (
@@ -31,6 +32,8 @@ var ModelRatio = map[string]float64{
 	"gpt-4-turbo-preview":     5,    // $0.01 / 1K tokens
 	"gpt-4-turbo":             5,    // $0.01 / 1K tokens
 	"gpt-4-turbo-2024-04-09":  5,    // $0.01 / 1K tokens
+	"gpt-4o":                  2.5,  // $0.005 / 1K tokens
+	"gpt-4o-2024-05-13":       2.5,  // $0.005 / 1K tokens
 	"gpt-4-vision-preview":    5,    // $0.01 / 1K tokens
 	"gpt-3.5-turbo":           0.25, // $0.0005 / 1K tokens
 	"gpt-3.5-turbo-0301":      0.75,
@@ -137,6 +140,8 @@ var ModelRatio = map[string]float64{
 	"Baichuan2-Turbo-192k": 0.016 * RMB,
 	"Baichuan2-53B":        0.02 * RMB,
 	// https://api.minimax.chat/document/price
+	"abab6.5-chat":  0.03 * RMB,
+	"abab6.5s-chat": 0.01 * RMB,
 	"abab6-chat":    0.1 * RMB,
 	"abab5.5-chat":  0.015 * RMB,
 	"abab5.5s-chat": 0.005 * RMB,
@@ -147,11 +152,13 @@ var ModelRatio = map[string]float64{
 	"mistral-medium-latest": 2.7 / 1000 * USD,
 	"mistral-large-latest":  8.0 / 1000 * USD,
 	"mistral-embed":         0.1 / 1000 * USD,
-	// https://wow.groq.com/
-	"llama2-70b-4096":    0.7 / 1000 * USD,
-	"llama2-7b-2048":     0.1 / 1000 * USD,
+	// https://wow.groq.com/#:~:text=inquiries%C2%A0here.-,Model,-Current%20Speed
+	"llama3-70b-8192":    0.59 / 1000 * USD,
 	"mixtral-8x7b-32768": 0.27 / 1000 * USD,
+	"llama3-8b-8192":     0.05 / 1000 * USD,
 	"gemma-7b-it":        0.1 / 1000 * USD,
+	"llama2-70b-4096":    0.64 / 1000 * USD,
+	"llama2-7b-2048":     0.1 / 1000 * USD,
 	// https://platform.lingyiwanwu.com/docs#-计费单元
 	"yi-34b-chat-0205": 2.5 / 1000 * RMB,
 	"yi-34b-chat-200k": 12.0 / 1000 * RMB,
@@ -160,6 +167,20 @@ var ModelRatio = map[string]float64{
 	"step-1v-32k": 0.024 * RMB,
 	"step-1-32k":  0.024 * RMB,
 	"step-1-200k": 0.15 * RMB,
+	// https://cohere.com/pricing
+	"command":               0.5,
+	"command-nightly":       0.5,
+	"command-light":         0.5,
+	"command-light-nightly": 0.5,
+	"command-r":             0.5 / 1000 * USD,
+	"command-r-plus":        3.0 / 1000 * USD,
+	// https://platform.deepseek.com/api-docs/pricing/
+	"deepseek-chat":  1.0 / 1000 * RMB,
+	"deepseek-coder": 1.0 / 1000 * RMB,
+	// https://www.deepl.com/pro?cta=header-prices
+	"deepl-zh": 25.0 / 1000 * USD,
+	"deepl-en": 25.0 / 1000 * USD,
+	"deepl-ja": 25.0 / 1000 * USD,
 }
 
 var CompletionRatio = map[string]float64{}
@@ -215,6 +236,9 @@ func GetModelRatio(name string) float64 {
 	if strings.HasPrefix(name, "qwen-") && strings.HasSuffix(name, "-internet") {
 		name = strings.TrimSuffix(name, "-internet")
 	}
+	if strings.HasPrefix(name, "command-") && strings.HasSuffix(name, "-internet") {
+		name = strings.TrimSuffix(name, "-internet")
+	}
 	ratio, ok := ModelRatio[name]
 	if !ok {
 		ratio, ok = DefaultModelRatio[name]
@@ -258,7 +282,9 @@ func GetCompletionRatio(name string) float64 {
 		return 4.0 / 3.0
 	}
 	if strings.HasPrefix(name, "gpt-4") {
-		if strings.HasPrefix(name, "gpt-4-turbo") {
+		if strings.HasPrefix(name, "gpt-4-turbo") ||
+			strings.HasPrefix(name, "gpt-4o") ||
+			strings.HasSuffix(name, "preview") {
 			return 3
 		}
 		return 2
@@ -275,9 +301,22 @@ func GetCompletionRatio(name string) float64 {
 	if strings.HasPrefix(name, "gemini-") {
 		return 3
 	}
+	if strings.HasPrefix(name, "deepseek-") {
+		return 2
+	}
 	switch name {
 	case "llama2-70b-4096":
-		return 0.8 / 0.7
+		return 0.8 / 0.64
+	case "llama3-8b-8192":
+		return 2
+	case "llama3-70b-8192":
+		return 0.79 / 0.59
+	case "command", "command-light", "command-nightly", "command-light-nightly":
+		return 2
+	case "command-r":
+		return 3
+	case "command-r-plus":
+		return 5
 	}
 	return 1
 }

relay/channeltype/define.go

@@ -35,6 +35,12 @@ const (
 	LingYiWanWu
 	StepFun
 	AwsClaude
-
+	Coze
+	Cohere
+	DeepSeek
+	Cloudflare
+	DeepL
+	TogetherAI
+	Doubao
 	Dummy
 )

relay/channeltype/helper.go

@@ -27,6 +27,14 @@ func ToAPIType(channelType int) int {
 		apiType = apitype.Ollama
 	case AwsClaude:
 		apiType = apitype.AwsClaude
+	case Coze:
+		apiType = apitype.Coze
+	case Cohere:
+		apiType = apitype.Cohere
+	case Cloudflare:
+		apiType = apitype.Cloudflare
+	case DeepL:
+		apiType = apitype.DeepL
 	}
 
 	return apiType

relay/channeltype/url.go

@@ -24,7 +24,7 @@ var ChannelBaseURLs = []string{
 	"https://openrouter.ai/api",                 // 20
 	"https://api.aiproxy.io",                    // 21
 	"https://fastgpt.run/api/openapi",           // 22
-	"https://hunyuan.cloud.tencent.com",         // 23
+	"https://hunyuan.tencentcloudapi.com",       // 23
 	"https://generativelanguage.googleapis.com", // 24
 	"https://api.moonshot.cn",                   // 25
 	"https://api.baichuan-ai.com",               // 26
@@ -35,6 +35,13 @@ var ChannelBaseURLs = []string{
 	"https://api.lingyiwanwu.com",               // 31
 	"https://api.stepfun.com",                   // 32
 	"",                                          // 33
+	"https://api.coze.com",                      // 34
+	"https://api.cohere.ai",                     // 35
+	"https://api.deepseek.com",                  // 36
+	"https://api.cloudflare.com",                // 37
+	"https://api-free.deepl.com",                // 38
+	"https://api.together.xyz",                  // 39
+	"https://ark.cn-beijing.volces.com",         // 40
 }
 
 func init() {

relay/client/init.go

@@ -1,24 +0,0 @@
-package client
-
-import (
-	"github.com/songquanpeng/one-api/common/config"
-	"net/http"
-	"time"
-)
-
-var HTTPClient *http.Client
-var ImpatientHTTPClient *http.Client
-
-func init() {
-	if config.RelayTimeout == 0 {
-		HTTPClient = &http.Client{}
-	} else {
-		HTTPClient = &http.Client{
-			Timeout: time.Duration(config.RelayTimeout) * time.Second,
-		}
-	}
-
-	ImpatientHTTPClient = &http.Client{
-		Timeout: 5 * time.Second,
-	}
-}

relay/constant/common.go

@@ -1,3 +1,5 @@
 package constant
 
 var StopFinishReason = "stop"
+var StreamObject = "chat.completion.chunk"
+var NonStreamObject = "chat.completion"