Merge pull request #736 from xy3xy3/main

更正硅基流动的SenseVoiceSmall模型名字

.dockerignore

@@ -0,0 +1,7 @@
+.github
+.git
+*.md
+.vscode
+.gitignore
+Makefile
+docs

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

@@ -13,7 +13,7 @@ on:
 jobs:
   push_to_registries:
     name: Push Docker image to multiple registries
-    runs-on: ubuntu-latest
+    runs-on: self-hosted
     permissions:
       packages: write
       contents: read

.github/workflows/linux-release.yml

@@ -17,7 +17,7 @@ jobs:
           fetch-depth: 0
       - uses: actions/setup-node@v3
         with:
-          node-version: 16
+          node-version: 18
       - name: Build Frontend
         env:
           CI: ""
@@ -38,7 +38,7 @@ jobs:
       - name: Build Backend (arm64)
         run: |
           sudo apt-get update
-          sudo apt-get install gcc-aarch64-linux-gnu
+          DEBIAN_FRONTEND=noninteractive sudo apt-get install -y gcc-aarch64-linux-gnu
           CC=aarch64-linux-gnu-gcc CGO_ENABLED=1 GOOS=linux GOARCH=arm64 go build -ldflags "-s -w -X 'one-api/common.Version=$(git describe --tags)' -extldflags '-static'" -o one-api-arm64
 
       - name: Release

.github/workflows/macos-release.yml

@@ -9,7 +9,7 @@ on:
       - '!*-alpha*'
 jobs:
   release:
-    runs-on: macos-latest
+    runs-on: ubuntu-latest
     steps:
       - name: Checkout
         uses: actions/checkout@v3
@@ -17,7 +17,7 @@ jobs:
           fetch-depth: 0
       - uses: actions/setup-node@v3
         with:
-          node-version: 16
+          node-version: 18
       - name: Build Frontend
         env:
           CI: ""

.github/workflows/windows-release.yml

@@ -9,7 +9,7 @@ on:
       - '!*-alpha*'
 jobs:
   release:
-    runs-on: windows-latest
+    runs-on: ubuntu-latest
     defaults:
       run:
         shell: bash
@@ -20,7 +20,7 @@ jobs:
           fetch-depth: 0
       - uses: actions/setup-node@v3
         with:
-          node-version: 16
+          node-version: 18
       - name: Build Frontend
         env:
           CI: ""

.gitignore

@@ -8,4 +8,5 @@ build
 logs
 web/dist
 .env
-one-api
+one-api
+.DS_Store

Dockerfile

@@ -1,11 +1,11 @@
-FROM node:16 as builder
+FROM oven/bun:latest as builder
 
 WORKDIR /build
 COPY web/package.json .
-RUN npm install
+RUN bun install
 COPY ./web .
 COPY ./VERSION .
-RUN DISABLE_ESLINT_PLUGIN='true' VITE_REACT_APP_VERSION=$(cat VERSION) npm run build
+RUN DISABLE_ESLINT_PLUGIN='true' VITE_REACT_APP_VERSION=$(cat VERSION) bun run build
 
 FROM golang AS builder2
 
@@ -24,8 +24,8 @@ FROM alpine
 
 RUN apk update \
     && apk upgrade \
-    && apk add --no-cache ca-certificates tzdata \
-    && update-ca-certificates 2>/dev/null || true
+    && apk add --no-cache ca-certificates tzdata ffmpeg \
+    && update-ca-certificates
 
 COPY --from=builder2 /build/one-api /
 EXPOSE 3000

README.en.md

@@ -59,6 +59,10 @@
 13. 🎵 Added [Suno API](https://github.com/Suno-API/Suno-API) interface support, [Integration Guide](Suno.md)
 14. 🔄 Support for Rerank models, compatible with Cohere and Jina, can integrate with Dify, [Integration Guide](Rerank.md)
 15. ⚡ **[OpenAI Realtime API](https://platform.openai.com/docs/guides/realtime/integration)** - Support for OpenAI's Realtime API, including Azure channels
+16. 🧠 Support for setting reasoning effort through model name suffix:
+    - Add suffix `-high` to set high reasoning effort (e.g., `o3-mini-high`)
+    - Add suffix `-medium` to set medium reasoning effort
+    - Add suffix `-low` to set low reasoning effort
 
 ## Model Support
 This version additionally supports:
@@ -82,6 +86,9 @@ You can add custom models gpt-4-gizmo-* in channels. These are third-party model
 - `GEMINI_MODEL_MAP`: Specify Gemini model versions (v1/v1beta), format: "model:version", comma-separated
 - `COHERE_SAFETY_SETTING`: Cohere model [safety settings](https://docs.cohere.com/docs/safety-modes#overview), options: `NONE`, `CONTEXTUAL`, `STRICT`, default `NONE`
 - `GEMINI_VISION_MAX_IMAGE_NUM`: Gemini model maximum image number, default `16`, set to `-1` to disable
+- `MAX_FILE_DOWNLOAD_MB`: Maximum file download size in MB, default `20`
+- `CRYPTO_SECRET`: Encryption key for encrypting database content
+- `AZURE_DEFAULT_API_VERSION`: Azure channel default API version, if not specified in channel settings, use this version, default `2024-12-01-preview`
 
 ## Deployment
 > [!TIP]
@@ -92,6 +99,10 @@ You can add custom models gpt-4-gizmo-* in channels. These are third-party model
 > docker run --rm -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtower -cR
 > ```
 
+### Multi-Server Deployment
+- Must set `SESSION_SECRET` environment variable, otherwise login state will not be consistent across multiple servers.
+- If using a public Redis, must set `CRYPTO_SECRET` environment variable, otherwise Redis content will not be able to be obtained in multi-server deployment.
+
 ### Requirements
 - Local database (default): SQLite (Docker deployment must mount `/data` directory)
 - Remote database: MySQL >= 5.7.8, PgSQL >= 9.6

README.md

@@ -65,6 +65,10 @@
 14. 🔄 支持Rerank模型，目前兼容Cohere和Jina，可接入Dify，[对接文档](Rerank.md)
 15. ⚡ **[OpenAI Realtime API](https://platform.openai.com/docs/guides/realtime/integration)** - 支持OpenAI的Realtime API，支持Azure渠道
 16. 支持使用路由/chat2link 进入聊天界面
+17. 🧠 支持通过模型名称后缀设置 reasoning effort：
+    - 添加后缀 `-high` 设置为 high reasoning effort (例如: `o3-mini-high`)
+    - 添加后缀 `-medium` 设置为 medium reasoning effort (例如: `o3-mini-medium`)
+    - 添加后缀 `-low` 设置为 low reasoning effort (例如: `o3-mini-low`)
 
 ## 模型支持
 此版本额外支持以下模型：
@@ -85,9 +89,12 @@
 - `GET_MEDIA_TOKEN`：是否统计图片token，默认为 `true`，关闭后将不再在本地计算图片token，可能会导致和上游计费不同，此项覆盖 `GET_MEDIA_TOKEN_NOT_STREAM` 选项作用。
 - `GET_MEDIA_TOKEN_NOT_STREAM`：是否在非流（`stream=false`）情况下统计图片token，默认为 `true`。
 - `UPDATE_TASK`：是否更新异步任务（Midjourney、Suno），默认为 `true`，关闭后将不会更新任务进度。
-- `GEMINI_MODEL_MAP`：Gemini模型指定版本(v1/v1beta)，使用“模型:版本”指定，","分隔，例如：-e GEMINI_MODEL_MAP="gemini-1.5-pro-latest:v1beta,gemini-1.5-pro-001:v1beta"，为空则使用默认配置(v1beta)
-- `COHERE_SAFETY_SETTING`：Cohere模型[安全设置](https://docs.cohere.com/docs/safety-modes#overview)，可选值为 `NONE`, `CONTEXTUAL`，`STRICT`，默认为 `NONE`。
+- `GEMINI_MODEL_MAP`：Gemini模型指定版本(v1/v1beta)，使用"模型:版本"指定，","分隔，例如：-e GEMINI_MODEL_MAP="gemini-1.5-pro-latest:v1beta,gemini-1.5-pro-001:v1beta"，为空则使用默认配置(v1beta)
+- `COHERE_SAFETY_SETTING`：Cohere模型[安全设置](https://docs.cohere.com/docs/safety-modes#overview)，可选值为 `NONE`, `CONTEXTUAL`, `STRICT`，默认为 `NONE`。
 - `GEMINI_VISION_MAX_IMAGE_NUM`：Gemini模型最大图片数量，默认为 `16`，设置为 `-1` 则不限制。
+- `MAX_FILE_DOWNLOAD_MB`: 最大文件下载大小，单位 MB，默认为 `20`。
+- `CRYPTO_SECRET`：加密密钥，用于加密数据库内容。
+- `AZURE_DEFAULT_API_VERSION`：Azure渠道默认API版本，如果渠道设置中未指定API版本，则使用此版本，默认为 `2024-12-01-preview`
 ## 部署
 > [!TIP]
 > 最新版Docker镜像：`calciumion/new-api:latest`  
@@ -97,6 +104,10 @@
 > docker run --rm -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtower -cR
 > ```
 
+### 多机部署
+- 必须设置环境变量 `SESSION_SECRET`，否则会导致多机部署时登录状态不一致。
+- 如果公用Redis，必须设置 `CRYPTO_SECRET`，否则会导致多机部署时Redis内容无法获取。
+
 ### 部署要求
 - 本地数据库（默认）：SQLite（Docker 部署默认使用 SQLite，必须挂载 `/data` 目录到宿主机）
 - 远程数据库：MySQL 版本 >= 5.7.8，PgSQL 版本 >= 9.6
@@ -153,15 +164,14 @@ docker run --name new-api -d --restart always -p 3000:3000 -e SQL_DSN="root:1234
 [对接文档](Suno.md)
 
 ## 界面截图
-![796df8d287b7b7bd7853b2497e7df511](https://github.com/user-attachments/assets/255b5e97-2d3a-4434-b4fa-e922ad88ff5a)
+![image](https://github.com/user-attachments/assets/a0dcd349-5df8-4dc8-9acf-ca272b239919)
+
+
+![image](https://github.com/user-attachments/assets/c7d0f7e1-729c-43e2-ac7c-2cb73b0afc8e)
 
-![image](https://github.com/Calcium-Ion/new-api/assets/61247483/ad0e7aae-0203-471c-9716-2d83768927d4)
 ![image](https://github.com/user-attachments/assets/29f81de5-33fc-4fc5-a5ff-f9b54b653c7c)
 
-![image](https://github.com/Calcium-Ion/new-api/assets/61247483/3ca0b282-00ff-4c96-bf9d-e29ef615c605)
-夜间模式  
-![image](https://github.com/Calcium-Ion/new-api/assets/61247483/1c66b593-bb9e-4757-9720-ff2759539242)
-![image](https://github.com/Calcium-Ion/new-api/assets/61247483/af9a07ee-5101-4b3d-8bd9-ae21a4fd7e9e)
+![image](https://github.com/user-attachments/assets/4fa53e18-d2c5-477a-9b26-b86e44c71e35)
 
 ## 交流群
 <img src="https://github.com/user-attachments/assets/9ca0bc82-e057-4230-a28d-9f198fa022e3" width="200">

common/constants.go

@@ -30,6 +30,7 @@ var DefaultCollapseSidebar = false // default value of collapse sidebar
 // Any options with "Secret", "Token" in its key won't be return by GetOptions
 
 var SessionSecret = uuid.New().String()
+var CryptoSecret = uuid.New().String()
 
 var OptionMap map[string]string
 var OptionMapRWMutex sync.RWMutex

common/crypto.go

@@ -1,6 +1,23 @@
 package common
 
-import "golang.org/x/crypto/bcrypt"
+import (
+	"crypto/hmac"
+	"crypto/sha256"
+	"encoding/hex"
+	"golang.org/x/crypto/bcrypt"
+)
+
+func GenerateHMACWithKey(key []byte, data string) string {
+	h := hmac.New(sha256.New, key)
+	h.Write([]byte(data))
+	return hex.EncodeToString(h.Sum(nil))
+}
+
+func GenerateHMAC(data string) string {
+	h := hmac.New(sha256.New, []byte(CryptoSecret))
+	h.Write([]byte(data))
+	return hex.EncodeToString(h.Sum(nil))
+}
 
 func Password2Hash(password string) (string, error) {
 	passwordBytes := []byte(password)

common/init.go

@@ -22,7 +22,7 @@ func printHelp() {
 	fmt.Println("Usage: one-api [--port <port>] [--log-dir <log directory>] [--version] [--help]")
 }
 
-func init() {
+func LoadEnv() {
 	flag.Parse()
 
 	if *PrintVersion {
@@ -45,6 +45,11 @@ func init() {
 			SessionSecret = ss
 		}
 	}
+	if os.Getenv("CRYPTO_SECRET") != "" {
+		CryptoSecret = os.Getenv("CRYPTO_SECRET")
+	} else {
+		CryptoSecret = SessionSecret
+	}
 	if os.Getenv("SQLITE_PATH") != "" {
 		SQLitePath = os.Getenv("SQLITE_PATH")
 	}

common/logger.go

@@ -36,7 +36,7 @@ func SetupLogger() {
 			setupLogLock.Unlock()
 			setupLogWorking = false
 		}()
-		logPath := filepath.Join(*LogDir, fmt.Sprintf("oneapi-%s.log", time.Now().Format("20060102")))
+		logPath := filepath.Join(*LogDir, fmt.Sprintf("oneapi-%s.log", time.Now().Format("20060102150405")))
 		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")

common/model-ratio.go

@@ -32,30 +32,42 @@ var defaultModelRatio = map[string]float64{
 	"gpt-4-0613": 15,
 	"gpt-4-32k":  30,
 	//"gpt-4-32k-0314":               30, //deprecated
-	"gpt-4-32k-0613":                  30,
-	"gpt-4-1106-preview":              5,    // $10 / 1M tokens
-	"gpt-4-0125-preview":              5,    // $10 / 1M tokens
-	"gpt-4-turbo-preview":             5,    // $10 / 1M tokens
-	"gpt-4-vision-preview":            5,    // $10 / 1M tokens
-	"gpt-4-1106-vision-preview":       5,    // $10 / 1M tokens
-	"chatgpt-4o-latest":               2.5,  // $5 / 1M tokens
-	"gpt-4o":                          1.25, // $2.5 / 1M tokens
-	"gpt-4o-audio-preview":            1.25, // $2.5 / 1M tokens
-	"gpt-4o-audio-preview-2024-10-01": 1.25, // $2.5 / 1M tokens
-	"gpt-4o-2024-05-13":               2.5,  // $5 / 1M tokens
-	"gpt-4o-2024-08-06":               1.25, // $2.5 / 1M tokens
-	"gpt-4o-2024-11-20":               1.25, // $2.5 / 1M tokens
-	"gpt-4o-realtime-preview":         2.5,
-	"o1":                              7.5,
-	"o1-2024-12-17":                   7.5,
-	"o1-preview":                      7.5,
-	"o1-preview-2024-09-12":           7.5,
-	"o1-mini":                         1.5,
-	"o1-mini-2024-09-12":              1.5,
-	"gpt-4o-mini":                     0.075,
-	"gpt-4o-mini-2024-07-18":          0.075,
-	"gpt-4-turbo":                     5, // $0.01 / 1K tokens
-	"gpt-4-turbo-2024-04-09":          5, // $0.01 / 1K tokens
+	"gpt-4-32k-0613":                          30,
+	"gpt-4-1106-preview":                      5,    // $10 / 1M tokens
+	"gpt-4-0125-preview":                      5,    // $10 / 1M tokens
+	"gpt-4-turbo-preview":                     5,    // $10 / 1M tokens
+	"gpt-4-vision-preview":                    5,    // $10 / 1M tokens
+	"gpt-4-1106-vision-preview":               5,    // $10 / 1M tokens
+	"chatgpt-4o-latest":                       2.5,  // $5 / 1M tokens
+	"gpt-4o":                                  1.25, // $2.5 / 1M tokens
+	"gpt-4o-audio-preview":                    1.25, // $2.5 / 1M tokens
+	"gpt-4o-audio-preview-2024-10-01":         1.25, // $2.5 / 1M tokens
+	"gpt-4o-2024-05-13":                       2.5,  // $5 / 1M tokens
+	"gpt-4o-2024-08-06":                       1.25, // $2.5 / 1M tokens
+	"gpt-4o-2024-11-20":                       1.25, // $2.5 / 1M tokens
+	"gpt-4o-realtime-preview":                 2.5,
+	"gpt-4o-realtime-preview-2024-10-01":      2.5,
+	"gpt-4o-realtime-preview-2024-12-17":      2.5,
+	"gpt-4o-mini-realtime-preview":            0.3,
+	"gpt-4o-mini-realtime-preview-2024-12-17": 0.3,
+	"o1":                        7.5,
+	"o1-2024-12-17":             7.5,
+	"o1-preview":                7.5,
+	"o1-preview-2024-09-12":     7.5,
+	"o1-mini":                   0.55,
+	"o1-mini-2024-09-12":        0.55,
+	"o3-mini":                   0.55,
+	"o3-mini-2025-01-31":        0.55,
+	"o3-mini-high":              0.55,
+	"o3-mini-2025-01-31-high":   0.55,
+	"o3-mini-low":               0.55,
+	"o3-mini-2025-01-31-low":    0.55,
+	"o3-mini-medium":            0.55,
+	"o3-mini-2025-01-31-medium": 0.55,
+	"gpt-4o-mini":               0.075,
+	"gpt-4o-mini-2024-07-18":    0.075,
+	"gpt-4-turbo":               5, // $0.01 / 1K tokens
+	"gpt-4-turbo-2024-04-09":    5, // $0.01 / 1K tokens
 	//"gpt-3.5-turbo-0301":           0.75, //deprecated
 	"gpt-3.5-turbo":          0.25,
 	"gpt-3.5-turbo-0613":     0.75,
@@ -179,8 +191,9 @@ var defaultModelRatio = map[string]float64{
 	"command-r-plus":         1.5,
 	"command-r-08-2024":      0.075,
 	"command-r-plus-08-2024": 1.25,
-	"deepseek-chat":          0.07,
-	"deepseek-coder":         0.07,
+	"deepseek-chat":          0.27 / 2,
+	"deepseek-coder":         0.27 / 2,
+	"deepseek-reasoner":      0.55 / 2, // 0.55 / 1k tokens
 	// Perplexity online 模型对搜索额外收费，有需要应自行调整，此处不计入搜索费用
 	"llama-3-sonar-small-32k-chat":   0.2 / 1000 * USD,
 	"llama-3-sonar-small-32k-online": 0.2 / 1000 * USD,
@@ -338,6 +351,12 @@ func UpdateCompletionRatioByJSONString(jsonStr string) error {
 }
 
 func GetCompletionRatio(name string) float64 {
+	if strings.Contains(name, "/") {
+		if ratio, ok := CompletionRatio[name]; ok {
+			return ratio
+		}
+	}
+	lowercaseName := strings.ToLower(name)
 	if strings.HasPrefix(name, "gpt-4-gizmo") {
 		name = "gpt-4-gizmo-*"
 	}
@@ -356,7 +375,7 @@ func GetCompletionRatio(name string) float64 {
 		}
 		return 2
 	}
-	if strings.HasPrefix(name, "o1-") {
+	if strings.HasPrefix(name, "o1") || strings.HasPrefix(name, "o3") {
 		return 4
 	}
 	if name == "chatgpt-4o-latest" {
@@ -397,11 +416,12 @@ func GetCompletionRatio(name string) float64 {
 		case "command-r-plus-08-2024":
 			return 4
 		default:
-			return 2
+			return 4
 		}
 	}
-	if strings.HasPrefix(name, "deepseek") {
-		return 2
+	// hint 只给官方上4倍率，由于开源模型供应商自行定价，不对其进行补全倍率进行强制对齐
+	if lowercaseName == "deepseek-chat" || lowercaseName == "deepseek-reasoner" {
+		return 4
 	}
 	if strings.HasPrefix(name, "ERNIE-Speed-") {
 		return 2
@@ -427,10 +447,23 @@ func GetCompletionRatio(name string) float64 {
 }
 
 func GetAudioRatio(name string) float64 {
-	if strings.HasPrefix(name, "gpt-4o-realtime") {
-		return 20
-	} else if strings.HasPrefix(name, "gpt-4o-audio") {
-		return 40
+	if strings.Contains(name, "-realtime") {
+		if strings.HasSuffix(name, "gpt-4o-realtime-preview-2024-12-17") {
+			return 8
+		} else if strings.Contains(name, "mini") {
+			return 10 / 0.6
+		} else {
+			return 20
+		}
+	}
+	if strings.Contains(name, "-audio") {
+		if strings.HasSuffix(name, "gpt-4o-audio-preview-2024-12-17") {
+			return 16
+		} else if strings.Contains(name, "mini") {
+			return 10 / 0.15
+		} else {
+			return 40
+		}
 	}
 	return 20
 }
@@ -438,6 +471,8 @@ func GetAudioRatio(name string) float64 {
 func GetAudioCompletionRatio(name string) float64 {
 	if strings.HasPrefix(name, "gpt-4o-realtime") {
 		return 2
+	} else if strings.HasPrefix(name, "gpt-4o-mini-realtime") {
+		return 2
 	}
 	return 2
 }

common/redis.go

@@ -2,9 +2,15 @@ package common
 
 import (
 	"context"
-	"github.com/go-redis/redis/v8"
+	"errors"
+	"fmt"
 	"os"
+	"reflect"
+	"strconv"
 	"time"
+
+	"github.com/go-redis/redis/v8"
+	"gorm.io/gorm"
 )
 
 var RDB *redis.Client
@@ -56,39 +62,167 @@ func RedisGet(key string) (string, error) {
 	return RDB.Get(ctx, key).Result()
 }
 
-func RedisExpire(key string, expiration time.Duration) error {
-	ctx := context.Background()
-	return RDB.Expire(ctx, key, expiration).Err()
-}
-
-func RedisGetEx(key string, expiration time.Duration) (string, error) {
-	ctx := context.Background()
-	return RDB.GetSet(ctx, key, expiration).Result()
-}
+//func RedisExpire(key string, expiration time.Duration) error {
+//	ctx := context.Background()
+//	return RDB.Expire(ctx, key, expiration).Err()
+//}
+//
+//func RedisGetEx(key string, expiration time.Duration) (string, error) {
+//	ctx := context.Background()
+//	return RDB.GetSet(ctx, key, expiration).Result()
+//}
 
 func RedisDel(key string) error {
 	ctx := context.Background()
 	return RDB.Del(ctx, key).Err()
 }
 
-func RedisDecrease(key string, value int64) error {
+func RedisHDelObj(key string) error {
+	ctx := context.Background()
+	return RDB.HDel(ctx, key).Err()
+}
 
+func RedisHSetObj(key string, obj interface{}, expiration time.Duration) error {
+	ctx := context.Background()
+
+	data := make(map[string]interface{})
+
+	// 使用反射遍历结构体字段
+	v := reflect.ValueOf(obj).Elem()
+	t := v.Type()
+	for i := 0; i < v.NumField(); i++ {
+		field := t.Field(i)
+		value := v.Field(i)
+
+		// Skip DeletedAt field
+		if field.Type.String() == "gorm.DeletedAt" {
+			continue
+		}
+
+		// 处理指针类型
+		if value.Kind() == reflect.Ptr {
+			if value.IsNil() {
+				data[field.Name] = ""
+				continue
+			}
+			value = value.Elem()
+		}
+
+		// 处理布尔类型
+		if value.Kind() == reflect.Bool {
+			data[field.Name] = strconv.FormatBool(value.Bool())
+			continue
+		}
+
+		// 其他类型直接转换为字符串
+		data[field.Name] = fmt.Sprintf("%v", value.Interface())
+	}
+
+	txn := RDB.TxPipeline()
+	txn.HSet(ctx, key, data)
+	txn.Expire(ctx, key, expiration)
+
+	_, err := txn.Exec(ctx)
+	if err != nil {
+		return fmt.Errorf("failed to execute transaction: %w", err)
+	}
+	return nil
+}
+
+func RedisHGetObj(key string, obj interface{}) error {
+	ctx := context.Background()
+
+	result, err := RDB.HGetAll(ctx, key).Result()
+	if err != nil {
+		return fmt.Errorf("failed to load hash from Redis: %w", err)
+	}
+
+	if len(result) == 0 {
+		return fmt.Errorf("key %s not found in Redis", key)
+	}
+
+	// Handle both pointer and non-pointer values
+	val := reflect.ValueOf(obj)
+	if val.Kind() != reflect.Ptr {
+		return fmt.Errorf("obj must be a pointer to a struct, got %T", obj)
+	}
+
+	v := val.Elem()
+	if v.Kind() != reflect.Struct {
+		return fmt.Errorf("obj must be a pointer to a struct, got pointer to %T", v.Interface())
+	}
+
+	t := v.Type()
+	for i := 0; i < v.NumField(); i++ {
+		field := t.Field(i)
+		fieldName := field.Name
+		if value, ok := result[fieldName]; ok {
+			fieldValue := v.Field(i)
+
+			// Handle pointer types
+			if fieldValue.Kind() == reflect.Ptr {
+				if value == "" {
+					continue
+				}
+				if fieldValue.IsNil() {
+					fieldValue.Set(reflect.New(fieldValue.Type().Elem()))
+				}
+				fieldValue = fieldValue.Elem()
+			}
+
+			// Enhanced type handling for Token struct
+			switch fieldValue.Kind() {
+			case reflect.String:
+				fieldValue.SetString(value)
+			case reflect.Int, reflect.Int64:
+				intValue, err := strconv.ParseInt(value, 10, 64)
+				if err != nil {
+					return fmt.Errorf("failed to parse int field %s: %w", fieldName, err)
+				}
+				fieldValue.SetInt(intValue)
+			case reflect.Bool:
+				boolValue, err := strconv.ParseBool(value)
+				if err != nil {
+					return fmt.Errorf("failed to parse bool field %s: %w", fieldName, err)
+				}
+				fieldValue.SetBool(boolValue)
+			case reflect.Struct:
+				// Special handling for gorm.DeletedAt
+				if fieldValue.Type().String() == "gorm.DeletedAt" {
+					if value != "" {
+						timeValue, err := time.Parse(time.RFC3339, value)
+						if err != nil {
+							return fmt.Errorf("failed to parse DeletedAt field %s: %w", fieldName, err)
+						}
+						fieldValue.Set(reflect.ValueOf(gorm.DeletedAt{Time: timeValue, Valid: true}))
+					}
+				}
+			default:
+				return fmt.Errorf("unsupported field type: %s for field %s", fieldValue.Kind(), fieldName)
+			}
+		}
+	}
+
+	return nil
+}
+
+// RedisIncr Add this function to handle atomic increments
+func RedisIncr(key string, delta int64) error {
 	// 检查键的剩余生存时间
 	ttlCmd := RDB.TTL(context.Background(), key)
 	ttl, err := ttlCmd.Result()
-	if err != nil {
-		// 失败则尝试直接减少
-		return RDB.DecrBy(context.Background(), key, value).Err()
+	if err != nil && !errors.Is(err, redis.Nil) {
+		return fmt.Errorf("failed to get TTL: %w", err)
 	}
 
-	// 如果剩余生存时间大于0，则进行减少操作
+	// 只有在 key 存在且有 TTL 时才需要特殊处理
 	if ttl > 0 {
 		ctx := context.Background()
 		// 开始一个Redis事务
 		txn := RDB.TxPipeline()
 
 		// 减少余额
-		decrCmd := txn.DecrBy(ctx, key, value)
+		decrCmd := txn.IncrBy(ctx, key, delta)
 		if err := decrCmd.Err(); err != nil {
 			return err // 如果减少失败，则直接返回错误
 		}
@@ -99,8 +233,54 @@ func RedisDecrease(key string, value int64) error {
 		// 执行事务
 		_, err = txn.Exec(ctx)
 		return err
-	} else {
-		_ = RedisDel(key)
+	}
+	return nil
+}
+
+func RedisHIncrBy(key, field string, delta int64) error {
+	ttlCmd := RDB.TTL(context.Background(), key)
+	ttl, err := ttlCmd.Result()
+	if err != nil && !errors.Is(err, redis.Nil) {
+		return fmt.Errorf("failed to get TTL: %w", err)
+	}
+
+	if ttl > 0 {
+		ctx := context.Background()
+		txn := RDB.TxPipeline()
+
+		incrCmd := txn.HIncrBy(ctx, key, field, delta)
+		if err := incrCmd.Err(); err != nil {
+			return err
+		}
+
+		txn.Expire(ctx, key, ttl)
+
+		_, err = txn.Exec(ctx)
+		return err
+	}
+	return nil
+}
+
+func RedisHSetField(key, field string, value interface{}) error {
+	ttlCmd := RDB.TTL(context.Background(), key)
+	ttl, err := ttlCmd.Result()
+	if err != nil && !errors.Is(err, redis.Nil) {
+		return fmt.Errorf("failed to get TTL: %w", err)
+	}
+
+	if ttl > 0 {
+		ctx := context.Background()
+		txn := RDB.TxPipeline()
+
+		hsetCmd := txn.HSet(ctx, key, field, value)
+		if err := hsetCmd.Err(); err != nil {
+			return err
+		}
+
+		txn.Expire(ctx, key, ttl)
+
+		_, err = txn.Exec(ctx)
+		return err
 	}
 	return nil
 }

common/str.go

@@ -35,9 +35,7 @@ func StrToMap(str string) map[string]interface{} {
 	m := make(map[string]interface{})
 	err := json.Unmarshal([]byte(str), &m)
 	if err != nil {
-		return map[string]interface{}{
-			"result": str,
-		}
+		return nil
 	}
 	return m
 }

common/user_groups.go

@@ -1,46 +0,0 @@
-package common
-
-import (
-	"encoding/json"
-)
-
-var UserUsableGroups = map[string]string{
-	"default": "默认分组",
-	"vip":     "vip分组",
-}
-
-func UserUsableGroups2JSONString() string {
-	jsonBytes, err := json.Marshal(UserUsableGroups)
-	if err != nil {
-		SysError("error marshalling user groups: " + err.Error())
-	}
-	return string(jsonBytes)
-}
-
-func UpdateUserUsableGroupsByJSONString(jsonStr string) error {
-	UserUsableGroups = make(map[string]string)
-	return json.Unmarshal([]byte(jsonStr), &UserUsableGroups)
-}
-
-func GetUserUsableGroups(userGroup string) map[string]string {
-	if userGroup == "" {
-		// 如果userGroup为空，返回UserUsableGroups
-		return UserUsableGroups
-	}
-	// 如果userGroup不在UserUsableGroups中，返回UserUsableGroups + userGroup
-	if _, ok := UserUsableGroups[userGroup]; !ok {
-		appendUserUsableGroups := make(map[string]string)
-		for k, v := range UserUsableGroups {
-			appendUserUsableGroups[k] = v
-		}
-		appendUserUsableGroups[userGroup] = "用户分组"
-		return appendUserUsableGroups
-	}
-	// 如果userGroup在UserUsableGroups中，返回UserUsableGroups
-	return UserUsableGroups
-}
-
-func GroupInUserUsableGroups(groupName string) bool {
-	_, ok := UserUsableGroups[groupName]
-	return ok
-}

common/utils.go

@@ -1,14 +1,19 @@
 package common
 
 import (
+	"bytes"
+	"context"
 	crand "crypto/rand"
 	"encoding/base64"
 	"fmt"
+	"github.com/pkg/errors"
 	"html/template"
+	"io"
 	"log"
 	"math/big"
 	"math/rand"
 	"net"
+	"os"
 	"os/exec"
 	"runtime"
 	"strconv"
@@ -207,3 +212,31 @@ func RandomSleep() {
 	// Sleep for 0-3000 ms
 	time.Sleep(time.Duration(rand.Intn(3000)) * time.Millisecond)
 }
+
+// SaveTmpFile saves data to a temporary file. The filename would be apppended with a random string.
+func SaveTmpFile(filename string, data io.Reader) (string, error) {
+	f, err := os.CreateTemp(os.TempDir(), filename)
+	if err != nil {
+		return "", errors.Wrapf(err, "failed to create temporary file %s", filename)
+	}
+	defer f.Close()
+
+	_, err = io.Copy(f, data)
+	if err != nil {
+		return "", errors.Wrapf(err, "failed to copy data to temporary file %s", filename)
+	}
+
+	return f.Name(), nil
+}
+
+// GetAudioDuration returns the duration of an audio file in seconds.
+func GetAudioDuration(ctx context.Context, filename string) (float64, error) {
+	// ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 {{input}}
+	c := exec.CommandContext(ctx, "ffprobe", "-v", "error", "-show_entries", "format=duration", "-of", "default=noprint_wrappers=1:nokey=1", filename)
+	output, err := c.Output()
+	if err != nil {
+		return 0, errors.Wrap(err, "failed to get audio duration")
+	}
+
+	return strconv.ParseFloat(string(bytes.TrimSpace(output)), 64)
+}

constant/cache_key.go

@@ -0,0 +1,23 @@
+package constant
+
+import "one-api/common"
+
+var (
+	TokenCacheSeconds         = common.SyncFrequency
+	UserId2GroupCacheSeconds  = common.SyncFrequency
+	UserId2QuotaCacheSeconds  = common.SyncFrequency
+	UserId2StatusCacheSeconds = common.SyncFrequency
+)
+
+// Cache keys
+const (
+	UserGroupKeyFmt    = "user_group:%d"
+	UserQuotaKeyFmt    = "user_quota:%d"
+	UserEnabledKeyFmt  = "user_enabled:%d"
+	UserUsernameKeyFmt = "user_name:%d"
+)
+
+const (
+	TokenFiledRemainQuota = "RemainQuota"
+	TokenFieldGroup       = "Group"
+)

constant/channel_setting.go

@@ -1,5 +1,6 @@
 package constant
 
 var (
-	ForceFormat = "force_format" // ForceFormat 强制格式化为OpenAI格式
+	ForceFormat        = "force_format" // ForceFormat 强制格式化为OpenAI格式
+	ChanelSettingProxy = "proxy"        // Proxy 代理
 )

constant/env.go

@@ -10,6 +10,8 @@ import (
 var StreamingTimeout = common.GetEnvOrDefault("STREAMING_TIMEOUT", 60)
 var DifyDebug = common.GetEnvOrDefaultBool("DIFY_DEBUG", true)
 
+var MaxFileDownloadMB = common.GetEnvOrDefault("MAX_FILE_DOWNLOAD_MB", 20)
+
 // ForceStreamOption 覆盖请求参数，强制返回usage信息
 var ForceStreamOption = common.GetEnvOrDefaultBool("FORCE_STREAM_OPTION", true)
 
@@ -19,6 +21,8 @@ var GetMediaTokenNotStream = common.GetEnvOrDefaultBool("GET_MEDIA_TOKEN_NOT_STR
 
 var UpdateTask = common.GetEnvOrDefaultBool("UPDATE_TASK", true)
 
+var AzureDefaultAPIVersion = common.GetEnvOrDefaultString("AZURE_DEFAULT_API_VERSION", "2024-12-01-preview")
+
 var GeminiModelMap = map[string]string{
 	"gemini-1.0-pro": "v1",
 }

constant/finish_reason.go

@@ -1,6 +1,9 @@
 package constant
 
 var (
-	FinishReasonStop      = "stop"
-	FinishReasonToolCalls = "tool_calls"
+	FinishReasonStop          = "stop"
+	FinishReasonToolCalls     = "tool_calls"
+	FinishReasonLength        = "length"
+	FinishReasonFunctionCall  = "function_call"
+	FinishReasonContentFilter = "content_filter"
 )

controller/billing.go

@@ -21,7 +21,7 @@ func GetSubscription(c *gin.Context) {
 		usedQuota = token.UsedQuota
 	} else {
 		userId := c.GetInt("id")
-		remainQuota, err = model.GetUserQuota(userId)
+		remainQuota, err = model.GetUserQuota(userId, false)
 		usedQuota, err = model.GetUserUsedQuota(userId)
 	}
 	if expiredTime <= 0 {

controller/channel-billing.go

@@ -78,6 +78,36 @@ type APGC2DGPTUsageResponse struct {
 	TotalUsed      float64 `json:"total_used"`
 }
 
+type SiliconFlowUsageResponse struct {
+	Code    int    `json:"code"`
+	Message string `json:"message"`
+	Status  bool   `json:"status"`
+	Data    struct {
+		ID            string `json:"id"`
+		Name          string `json:"name"`
+		Image         string `json:"image"`
+		Email         string `json:"email"`
+		IsAdmin       bool   `json:"isAdmin"`
+		Balance       string `json:"balance"`
+		Status        string `json:"status"`
+		Introduction  string `json:"introduction"`
+		Role          string `json:"role"`
+		ChargeBalance string `json:"chargeBalance"`
+		TotalBalance  string `json:"totalBalance"`
+		Category      string `json:"category"`
+	} `json:"data"`
+}
+
+type DeepSeekUsageResponse struct {
+	IsAvailable  bool `json:"is_available"`
+	BalanceInfos []struct {
+		Currency        string `json:"currency"`
+		TotalBalance    string `json:"total_balance"`
+		GrantedBalance  string `json:"granted_balance"`
+		ToppedUpBalance string `json:"topped_up_balance"`
+	} `json:"balance_infos"`
+}
+
 // GetAuthHeader get auth header
 func GetAuthHeader(token string) http.Header {
 	h := http.Header{}
@@ -185,6 +215,57 @@ func updateChannelAPI2GPTBalance(channel *model.Channel) (float64, error) {
 	return response.TotalRemaining, nil
 }
 
+func updateChannelSiliconFlowBalance(channel *model.Channel) (float64, error) {
+	url := "https://api.siliconflow.cn/v1/user/info"
+	body, err := GetResponseBody("GET", url, channel, GetAuthHeader(channel.Key))
+	if err != nil {
+		return 0, err
+	}
+	response := SiliconFlowUsageResponse{}
+	err = json.Unmarshal(body, &response)
+	if err != nil {
+		return 0, err
+	}
+	if response.Code != 20000 {
+		return 0, fmt.Errorf("code: %d, message: %s", response.Code, response.Message)
+	}
+	balance, err := strconv.ParseFloat(response.Data.TotalBalance, 64)
+	if err != nil {
+		return 0, err
+	}
+	channel.UpdateBalance(balance)
+	return balance, nil
+}
+
+func updateChannelDeepSeekBalance(channel *model.Channel) (float64, error) {
+	url := "https://api.deepseek.com/user/balance"
+	body, err := GetResponseBody("GET", url, channel, GetAuthHeader(channel.Key))
+	if err != nil {
+		return 0, err
+	}
+	response := DeepSeekUsageResponse{}
+	err = json.Unmarshal(body, &response)
+	if err != nil {
+		return 0, err
+	}
+	index := -1
+	for i, balanceInfo := range response.BalanceInfos {
+		if balanceInfo.Currency == "CNY" {
+			index = i
+			break
+		}
+	}
+	if index == -1 {
+		return 0, errors.New("currency CNY not found")
+	}
+	balance, err := strconv.ParseFloat(response.BalanceInfos[index].TotalBalance, 64)
+	if err != nil {
+		return 0, err
+	}
+	channel.UpdateBalance(balance)
+	return balance, nil
+}
+
 func updateChannelAIGC2DBalance(channel *model.Channel) (float64, error) {
 	url := "https://api.aigc2d.com/dashboard/billing/credit_grants"
 	body, err := GetResponseBody("GET", url, channel, GetAuthHeader(channel.Key))
@@ -222,6 +303,10 @@ func updateChannelBalance(channel *model.Channel) (float64, error) {
 		return updateChannelAPI2GPTBalance(channel)
 	case common.ChannelTypeAIGC2D:
 		return updateChannelAIGC2DBalance(channel)
+	case common.ChannelTypeSiliconFlow:
+		return updateChannelSiliconFlowBalance(channel)
+	case common.ChannelTypeDeepSeek:
+		return updateChannelDeepSeekBalance(channel)
 	default:
 		return 0, errors.New("尚未实现")
 	}
@@ -300,9 +385,9 @@ func updateAllChannelsBalance() error {
 			continue
 		}
 		// TODO: support Azure
-		if channel.Type != common.ChannelTypeOpenAI && channel.Type != common.ChannelTypeCustom {
-			continue
-		}
+		//if channel.Type != common.ChannelTypeOpenAI && channel.Type != common.ChannelTypeCustom {
+		//	continue
+		//}
 		balance, err := updateChannelBalance(channel)
 		if err != nil {
 			continue

controller/channel-test.go

@@ -5,7 +5,6 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/bytedance/gopkg/util/gopool"
 	"io"
 	"math"
 	"net/http"
@@ -24,6 +23,8 @@ import (
 	"sync"
 	"time"
 
+	"github.com/bytedance/gopkg/util/gopool"
+
 	"github.com/gin-gonic/gin"
 )
 
@@ -32,6 +33,9 @@ func testChannel(channel *model.Channel, testModel string) (err error, openAIErr
 	if channel.Type == common.ChannelTypeMidjourney {
 		return errors.New("midjourney channel test is not supported"), nil
 	}
+	if channel.Type == common.ChannelTypeMidjourneyPlus {
+		return errors.New("midjourney plus channel test is not supported!!!"), nil
+	}
 	if channel.Type == common.ChannelTypeSunoAPI {
 		return errors.New("suno channel test is not supported"), nil
 	}
@@ -152,10 +156,10 @@ func buildTestRequest(model string) *dto.GeneralOpenAIRequest {
 		Model:  "", // this will be set later
 		Stream: false,
 	}
-	if strings.HasPrefix(model, "o1-") {
-		testRequest.MaxCompletionTokens = 1
+	if strings.HasPrefix(model, "o1") || strings.HasPrefix(model, "o3") {
+		testRequest.MaxCompletionTokens = 10
 	} else if strings.HasPrefix(model, "gemini-2.0-flash-thinking") {
-		testRequest.MaxTokens = 2
+		testRequest.MaxTokens = 10
 	} else {
 		testRequest.MaxTokens = 1
 	}

controller/channel.go

@@ -274,6 +274,17 @@ func AddChannel(c *gin.Context) {
 		}
 		localChannel := channel
 		localChannel.Key = key
+		// Validate the length of the model name
+		models := strings.Split(localChannel.Models, ",")
+		for _, model := range models {
+			if len(model) > 255 {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": fmt.Sprintf("模型名称过长: %s", model),
+				})
+				return
+			}
+		}
 		channels = append(channels, localChannel)
 	}
 	err = model.BatchInsertChannels(channels)
@@ -419,7 +430,8 @@ func EditTagChannels(c *gin.Context) {
 }
 
 type ChannelBatch struct {
-	Ids []int `json:"ids"`
+	Ids []int   `json:"ids"`
+	Tag *string `json:"tag"`
 }
 
 func DeleteChannelBatch(c *gin.Context) {
@@ -498,6 +510,7 @@ func UpdateChannel(c *gin.Context) {
 func FetchModels(c *gin.Context) {
 	var req struct {
 		BaseURL string `json:"base_url"`
+		Type    int    `json:"type"`
 		Key     string `json:"key"`
 	}
 
@@ -511,7 +524,7 @@ func FetchModels(c *gin.Context) {
 
 	baseURL := req.BaseURL
 	if baseURL == "" {
-		baseURL = "https://api.openai.com"
+		baseURL = common.ChannelBaseURLs[req.Type]
 	}
 
 	client := &http.Client{}
@@ -526,7 +539,11 @@ func FetchModels(c *gin.Context) {
 		return
 	}
 
-	request.Header.Set("Authorization", "Bearer "+req.Key)
+	// remove line breaks and extra spaces.
+	key := strings.TrimSpace(req.Key)
+	// If the key contains a line break, only take the first part.
+	key = strings.Split(key, "\n")[0]
+	request.Header.Set("Authorization", "Bearer "+key)
 
 	response, err := client.Do(request)
 	if err != nil {
@@ -570,3 +587,29 @@ func FetchModels(c *gin.Context) {
 		"data":    models,
 	})
 }
+
+func BatchSetChannelTag(c *gin.Context) {
+	channelBatch := ChannelBatch{}
+	err := c.ShouldBindJSON(&channelBatch)
+	if err != nil || len(channelBatch.Ids) == 0 {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "参数错误",
+		})
+		return
+	}
+	err = model.BatchSetChannelTag(channelBatch.Ids, channelBatch.Tag)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "",
+		"data":    len(channelBatch.Ids),
+	})
+	return
+}

controller/group.go

@@ -3,13 +3,13 @@ package controller
 import (
 	"github.com/gin-gonic/gin"
 	"net/http"
-	"one-api/common"
 	"one-api/model"
+	"one-api/setting"
 )
 
 func GetGroups(c *gin.Context) {
 	groupNames := make([]string, 0)
-	for groupName, _ := range common.GroupRatio {
+	for groupName, _ := range setting.GetGroupRatioCopy() {
 		groupNames = append(groupNames, groupName)
 	}
 	c.JSON(http.StatusOK, gin.H{
@@ -20,15 +20,18 @@ func GetGroups(c *gin.Context) {
 }
 
 func GetUserGroups(c *gin.Context) {
-	usableGroups := make(map[string]string)
+	usableGroups := make(map[string]map[string]interface{})
 	userGroup := ""
 	userId := c.GetInt("id")
-	userGroup, _ = model.CacheGetUserGroup(userId)
-	for groupName, _ := range common.GroupRatio {
+	userGroup, _ = model.GetUserGroup(userId, false)
+	for groupName, ratio := range setting.GetGroupRatioCopy() {
 		// UserUsableGroups contains the groups that the user can use
-		userUsableGroups := common.GetUserUsableGroups(userGroup)
-		if _, ok := userUsableGroups[groupName]; ok {
-			usableGroups[groupName] = userUsableGroups[groupName]
+		userUsableGroups := setting.GetUserUsableGroups(userGroup)
+		if desc, ok := userUsableGroups[groupName]; ok {
+			usableGroups[groupName] = map[string]interface{}{
+				"ratio": ratio,
+				"desc":  desc,
+			}
 		}
 	}
 	c.JSON(http.StatusOK, gin.H{

controller/misc.go

@@ -66,6 +66,7 @@ func GetStatus(c *gin.Context) {
 			"enable_online_topup":      setting.PayAddress != "" && setting.EpayId != "" && setting.EpayKey != "",
 			"mj_notify_enabled":        setting.MjNotifyEnabled,
 			"chats":                    setting.Chats,
+			"demo_site_enabled":        setting.DemoSiteEnabled,
 		},
 	})
 	return

controller/model.go

@@ -166,7 +166,7 @@ func ListModels(c *gin.Context) {
 		}
 	} else {
 		userId := c.GetInt("id")
-		userGroup, err := model.GetUserGroup(userId)
+		userGroup, err := model.GetUserGroup(userId, true)
 		if err != nil {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,

controller/option.go

@@ -5,6 +5,7 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/model"
+	"one-api/setting"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -83,7 +84,7 @@ func UpdateOption(c *gin.Context) {
 			return
 		}
 	case "GroupRatio":
-		err = common.CheckGroupRatio(option.Value)
+		err = setting.CheckGroupRatio(option.Value)
 		if err != nil {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,

controller/playground.go

@@ -0,0 +1,69 @@
+package controller
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"net/http"
+	"one-api/common"
+	"one-api/constant"
+	"one-api/dto"
+	"one-api/middleware"
+	"one-api/model"
+	"one-api/service"
+	"one-api/setting"
+	"time"
+)
+
+func Playground(c *gin.Context) {
+	var openaiErr *dto.OpenAIErrorWithStatusCode
+
+	defer func() {
+		if openaiErr != nil {
+			c.JSON(openaiErr.StatusCode, gin.H{
+				"error": openaiErr.Error,
+			})
+		}
+	}()
+
+	useAccessToken := c.GetBool("use_access_token")
+	if useAccessToken {
+		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("暂不支持使用 access token"), "access_token_not_supported", http.StatusBadRequest)
+		return
+	}
+
+	playgroundRequest := &dto.PlayGroundRequest{}
+	err := common.UnmarshalBodyReusable(c, playgroundRequest)
+	if err != nil {
+		openaiErr = service.OpenAIErrorWrapperLocal(err, "unmarshal_request_failed", http.StatusBadRequest)
+		return
+	}
+
+	if playgroundRequest.Model == "" {
+		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("请选择模型"), "model_required", http.StatusBadRequest)
+		return
+	}
+	c.Set("original_model", playgroundRequest.Model)
+	group := playgroundRequest.Group
+	userGroup := c.GetString("group")
+
+	if group == "" {
+		group = userGroup
+	} else {
+		if !setting.GroupInUserUsableGroups(group) && group != userGroup {
+			openaiErr = service.OpenAIErrorWrapperLocal(errors.New("无权访问该分组"), "group_not_allowed", http.StatusForbidden)
+			return
+		}
+		c.Set("group", group)
+	}
+	c.Set("token_name", "playground-"+group)
+	channel, err := model.CacheGetRandomSatisfiedChannel(group, playgroundRequest.Model, 0)
+	if err != nil {
+		message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", group, playgroundRequest.Model)
+		openaiErr = service.OpenAIErrorWrapperLocal(errors.New(message), "get_playground_channel_failed", http.StatusInternalServerError)
+		return
+	}
+	middleware.SetupContextForSelectedChannel(c, channel, playgroundRequest.Model)
+	c.Set(constant.ContextKeyRequestStartTime, time.Now())
+	Relay(c)
+}

controller/pricing.go

@@ -4,29 +4,28 @@ import (
 	"github.com/gin-gonic/gin"
 	"one-api/common"
 	"one-api/model"
+	"one-api/setting"
 )
 
 func GetPricing(c *gin.Context) {
 	pricing := model.GetPricing()
 	userId, exists := c.Get("id")
 	usableGroup := map[string]string{}
-	groupRatio := common.GroupRatio
+	groupRatio := map[string]float64{}
+	for s, f := range setting.GetGroupRatioCopy() {
+		groupRatio[s] = f
+	}
 	var group string
 	if exists {
-		user, err := model.GetChannelById(userId.(int), false)
-		if err != nil {
-			c.JSON(200, gin.H{
-				"success": false,
-				"message": err.Error(),
-			})
-			return
+		user, err := model.GetUserById(userId.(int), false)
+		if err == nil {
+			group = user.Group
 		}
-		group = user.Group
 	}
 
-	usableGroup = common.GetUserUsableGroups(group)
+	usableGroup = setting.GetUserUsableGroups(group)
 	// check groupRatio contains usableGroup
-	for group := range common.GroupRatio {
+	for group := range setting.GetGroupRatioCopy() {
 		if _, ok := usableGroup[group]; !ok {
 			delete(groupRatio, group)
 		}

controller/redemption.go

@@ -1,19 +1,24 @@
 package controller
 
 import (
-	"github.com/gin-gonic/gin"
 	"net/http"
 	"one-api/common"
 	"one-api/model"
 	"strconv"
+
+	"github.com/gin-gonic/gin"
 )
 
 func GetAllRedemptions(c *gin.Context) {
 	p, _ := strconv.Atoi(c.Query("p"))
+	pageSize, _ := strconv.Atoi(c.Query("page_size"))
 	if p < 0 {
 		p = 0
 	}
-	redemptions, err := model.GetAllRedemptions(p*common.ItemsPerPage, common.ItemsPerPage)
+	if pageSize < 1 {
+		pageSize = common.ItemsPerPage
+	}
+	redemptions, total, err := model.GetAllRedemptions((p-1)*pageSize, pageSize)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -24,14 +29,27 @@ func GetAllRedemptions(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data":    redemptions,
+		"data": gin.H{
+			"items":     redemptions,
+			"total":     total,
+			"page":      p,
+			"page_size": pageSize,
+		},
 	})
 	return
 }
 
 func SearchRedemptions(c *gin.Context) {
 	keyword := c.Query("keyword")
-	redemptions, err := model.SearchRedemptions(keyword)
+	p, _ := strconv.Atoi(c.Query("p"))
+	pageSize, _ := strconv.Atoi(c.Query("page_size"))
+	if p < 0 {
+		p = 0
+	}
+	if pageSize < 1 {
+		pageSize = common.ItemsPerPage
+	}
+	redemptions, total, err := model.SearchRedemptions(keyword, (p-1)*pageSize, pageSize)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -42,7 +60,12 @@ func SearchRedemptions(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data":    redemptions,
+		"data": gin.H{
+			"items":     redemptions,
+			"total":     total,
+			"page":      p,
+			"page_size": pageSize,
+		},
 	})
 	return
 }

controller/relay.go

@@ -48,58 +48,6 @@ func wsHandler(c *gin.Context, ws *websocket.Conn, relayMode int) *dto.OpenAIErr
 	return err
 }
 
-func Playground(c *gin.Context) {
-	var openaiErr *dto.OpenAIErrorWithStatusCode
-
-	defer func() {
-		if openaiErr != nil {
-			c.JSON(openaiErr.StatusCode, gin.H{
-				"error": openaiErr.Error,
-			})
-		}
-	}()
-
-	useAccessToken := c.GetBool("use_access_token")
-	if useAccessToken {
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("暂不支持使用 access token"), "access_token_not_supported", http.StatusBadRequest)
-		return
-	}
-
-	playgroundRequest := &dto.PlayGroundRequest{}
-	err := common.UnmarshalBodyReusable(c, playgroundRequest)
-	if err != nil {
-		openaiErr = service.OpenAIErrorWrapperLocal(err, "unmarshal_request_failed", http.StatusBadRequest)
-		return
-	}
-
-	if playgroundRequest.Model == "" {
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New("请选择模型"), "model_required", http.StatusBadRequest)
-		return
-	}
-	c.Set("original_model", playgroundRequest.Model)
-	group := playgroundRequest.Group
-	userGroup := c.GetString("group")
-
-	if group == "" {
-		group = userGroup
-	} else {
-		if !common.GroupInUserUsableGroups(group) && group != userGroup {
-			openaiErr = service.OpenAIErrorWrapperLocal(errors.New("无权访问该分组"), "group_not_allowed", http.StatusForbidden)
-			return
-		}
-		c.Set("group", group)
-	}
-	c.Set("token_name", "playground-"+group)
-	channel, err := model.CacheGetRandomSatisfiedChannel(group, playgroundRequest.Model, 0)
-	if err != nil {
-		message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", group, playgroundRequest.Model)
-		openaiErr = service.OpenAIErrorWrapperLocal(errors.New(message), "get_playground_channel_failed", http.StatusInternalServerError)
-		return
-	}
-	middleware.SetupContextForSelectedChannel(c, channel, playgroundRequest.Model)
-	Relay(c)
-}
-
 func Relay(c *gin.Context) {
 	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
 	requestId := c.GetString(common.RequestIdKey)

controller/task.go

@@ -153,7 +153,7 @@ func updateSunoTaskAll(ctx context.Context, channelId int, taskIds []string, tas
 		if responseItem.FailReason != "" || task.Status == model.TaskStatusFailure {
 			common.LogInfo(ctx, task.TaskID+" 构建失败，"+task.FailReason)
 			task.Progress = "100%"
-			err = model.CacheUpdateUserQuota(task.UserId)
+			//err = model.CacheUpdateUserQuota(task.UserId) ?
 			if err != nil {
 				common.LogError(ctx, "error update user quota cache: "+err.Error())
 			} else {

controller/topup.go

@@ -75,7 +75,7 @@ func RequestEpay(c *gin.Context) {
 	}
 
 	id := c.GetInt("id")
-	group, err := model.CacheGetUserGroup(id)
+	group, err := model.GetUserGroup(id, true)
 	if err != nil {
 		c.JSON(200, gin.H{"message": "error", "data": "获取用户分组失败"})
 		return
@@ -236,7 +236,7 @@ func RequestAmount(c *gin.Context) {
 		return
 	}
 	id := c.GetInt("id")
-	group, err := model.CacheGetUserGroup(id)
+	group, err := model.GetUserGroup(id, true)
 	if err != nil {
 		c.JSON(200, gin.H{"message": "error", "data": "获取用户分组失败"})
 		return

controller/user.go

@@ -6,13 +6,15 @@ import (
 	"net/http"
 	"one-api/common"
 	"one-api/model"
+	"one-api/setting"
 	"strconv"
 	"strings"
 	"sync"
 
+	"one-api/constant"
+
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
-	"one-api/constant"
 )
 
 type LoginRequest struct {
@@ -241,10 +243,14 @@ func Register(c *gin.Context) {
 
 func GetAllUsers(c *gin.Context) {
 	p, _ := strconv.Atoi(c.Query("p"))
-	if p < 0 {
-		p = 0
+	pageSize, _ := strconv.Atoi(c.Query("page_size"))
+	if p < 1 {
+		p = 1
 	}
-	users, err := model.GetAllUsers(p*common.ItemsPerPage, common.ItemsPerPage)
+	if pageSize < 0 {
+		pageSize = common.ItemsPerPage
+	}
+	users, total, err := model.GetAllUsers((p-1)*pageSize, pageSize)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -255,7 +261,12 @@ func GetAllUsers(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data":    users,
+		"data": gin.H{
+			"items":     users,
+			"total":     total,
+			"page":      p,
+			"page_size": pageSize,
+		},
 	})
 	return
 }
@@ -263,7 +274,16 @@ func GetAllUsers(c *gin.Context) {
 func SearchUsers(c *gin.Context) {
 	keyword := c.Query("keyword")
 	group := c.Query("group")
-	users, err := model.SearchUsers(keyword, group)
+	p, _ := strconv.Atoi(c.Query("p"))
+	pageSize, _ := strconv.Atoi(c.Query("page_size"))
+	if p < 1 {
+		p = 1
+	}
+	if pageSize < 0 {
+		pageSize = common.ItemsPerPage
+	}
+	startIdx := (p - 1) * pageSize
+	users, total, err := model.SearchUsers(keyword, group, startIdx, pageSize)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -274,7 +294,12 @@ func SearchUsers(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
-		"data":    users,
+		"data": gin.H{
+			"items":     users,
+			"total":     total,
+			"page":      p,
+			"page_size": pageSize,
+		},
 	})
 	return
 }
@@ -454,7 +479,15 @@ func GetUserModels(c *gin.Context) {
 		})
 		return
 	}
-	models := model.GetGroupModels(user.Group)
+	groups := setting.GetUserUsableGroups(user.Group)
+	var models []string
+	for group := range groups {
+		for _, g := range model.GetGroupModels(group) {
+			if !common.StringsContains(models, g) {
+				models = append(models, g)
+			}
+		}
+	}
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",

docs/api/api_auth.md

@@ -0,0 +1,53 @@
+# API 鉴权文档
+
+## 认证方式
+
+### Access Token
+
+对于需要鉴权的 API 接口，必须同时提供以下两个请求头来进行 Access Token 认证：
+
+1. **请求头中的 `Authorization` 字段**
+
+    将 Access Token 放置于 HTTP 请求头部的 `Authorization` 字段中，格式如下：
+
+    ```
+    Authorization: <your_access_token>
+    ```
+
+    其中 `<your_access_token>` 需要替换为实际的 Access Token 值。
+
+2. **请求头中的 `New-Api-User` 字段**
+
+    将用户 ID 放置于 HTTP 请求头部的 `New-Api-User` 字段中，格式如下：
+
+    ```
+    New-Api-User: <your_user_id>
+    ```
+
+    其中 `<your_user_id>` 需要替换为实际的用户 ID。
+
+**注意：**
+
+*   **必须同时提供 `Authorization` 和 `New-Api-User` 两个请求头才能通过鉴权。**
+*   如果只提供其中一个请求头，或者两个请求头都未提供，则会返回 `401 Unauthorized` 错误。
+*   如果 `Authorization` 中的 Access Token 无效，则会返回 `401 Unauthorized` 错误，并提示“无权进行此操作，access token 无效”。
+*   如果 `New-Api-User` 中的用户 ID 与 Access Token 不匹配，则会返回 `401 Unauthorized` 错误，并提示“无权进行此操作，与登录用户不匹配，请重新登录”。
+*   如果没有提供 `New-Api-User` 请求头，则会返回 `401 Unauthorized` 错误，并提示“无权进行此操作，未提供 New-Api-User”。
+*   如果 `New-Api-User` 请求头格式错误，则会返回 `401 Unauthorized` 错误，并提示“无权进行此操作，New-Api-User 格式错误”。
+*   如果用户已被禁用，则会返回 `403 Forbidden` 错误，并提示“用户已被封禁”。
+*   如果用户权限不足，则会返回 `403 Forbidden` 错误，并提示“无权进行此操作，权限不足”。
+*   如果用户信息无效，则会返回 `403 Forbidden` 错误，并提示“无权进行此操作，用户信息无效”。
+
+## Curl 示例
+
+假设您的 Access Token 为 `access_token`，用户 ID 为 `123`，要访问的 API 接口为 `/api/user/self`，则可以使用以下 curl 命令：
+
+```bash
+curl -X GET \
+  -H "Authorization: access_token" \
+  -H "New-Api-User: 123" \
+  https://your-domain.com/api/user/self
+```
+
+请将 `access_token`、`123` 和 `https://your-domain.com` 替换为实际的值。
+

docs/channel/other_setting.md

@@ -0,0 +1,28 @@
+# 渠道而外设置说明
+
+该配置用于设置一些额外的渠道参数，可以通过 JSON 对象进行配置。主要包含以下两个设置项：
+
+1. force_format
+    - 用于标识是否对数据进行强制格式化为 OpenAI 格式
+    - 类型为布尔值，设置为 true 时启用强制格式化
+
+2. proxy
+    - 用于配置网络代理
+    - 类型为字符串，填写代理地址（例如 socks5 协议的代理地址）
+
+--------------------------------------------------------------
+
+## JSON 格式示例
+
+以下是一个示例配置，启用强制格式化并设置了代理地址：
+
+```json
+{
+    "force_format": true,
+    "proxy": "socks5://xxxxxxx"
+}
+```
+
+--------------------------------------------------------------
+
+通过调整上述 JSON 配置中的值，可以灵活控制渠道的额外行为，比如是否进行格式化以及使用特定的网络代理。

dto/file_data.go

@@ -0,0 +1,8 @@
+package dto
+
+type LocalFileData struct {
+	MimeType   string
+	Base64Data string
+	Url        string
+	Size       int64
+}

dto/openai_request.go

@@ -23,7 +23,7 @@ type GeneralOpenAIRequest struct {
 	MaxTokens           uint            `json:"max_tokens,omitempty"`
 	MaxCompletionTokens uint            `json:"max_completion_tokens,omitempty"`
 	ReasoningEffort     string          `json:"reasoning_effort,omitempty"`
-	Temperature         float64         `json:"temperature,omitempty"`
+	Temperature         *float64        `json:"temperature,omitempty"`
 	TopP                float64         `json:"top_p,omitempty"`
 	TopK                int             `json:"top_k,omitempty"`
 	Stop                any             `json:"stop,omitempty"`

dto/openai_response.go

@@ -86,6 +86,10 @@ type ToolCall struct {
 	Function FunctionCall `json:"function"`
 }
 
+func (c *ToolCall) SetIndex(i int) {
+	c.Index = &i
+}
+
 type FunctionCall struct {
 	Description string `json:"description,omitempty"`
 	Name        string `json:"name,omitempty"`

main.go

@@ -33,9 +33,11 @@ var indexPage []byte
 func main() {
 	err := godotenv.Load(".env")
 	if err != nil {
-		common.SysError("failed to load .env file: " + err.Error())
+		common.SysLog("Support for .env file is disabled")
 	}
 
+	common.LoadEnv()
+
 	common.SetupLogger()
 	common.SysLog("New API " + common.Version + " started")
 	if os.Getenv("GIN_MODE") != "debug" {
@@ -80,9 +82,6 @@ func main() {
 		common.SysError(fmt.Sprintf("sync frequency: %d seconds", common.SyncFrequency))
 		model.InitChannelCache()
 	}
-	if common.RedisEnabled {
-		go model.SyncTokenCache(common.SyncFrequency)
-	}
 	if common.MemoryCacheEnabled {
 		go model.SyncOptions(common.SyncFrequency)
 		go model.SyncChannelCache(common.SyncFrequency)

middleware/auth.go

@@ -64,35 +64,33 @@ func authHelper(c *gin.Context, minRole int) {
 			return
 		}
 	}
-	if !useAccessToken {
-		// get header New-Api-User
-		apiUserIdStr := c.Request.Header.Get("New-Api-User")
-		if apiUserIdStr == "" {
-			c.JSON(http.StatusUnauthorized, gin.H{
-				"success": false,
-				"message": "无权进行此操作，请刷新页面或清空缓存后重试",
-			})
-			c.Abort()
-			return
-		}
-		apiUserId, err := strconv.Atoi(apiUserIdStr)
-		if err != nil {
-			c.JSON(http.StatusUnauthorized, gin.H{
-				"success": false,
-				"message": "无权进行此操作，登录信息无效，请重新登录",
-			})
-			c.Abort()
-			return
+	// get header New-Api-User
+	apiUserIdStr := c.Request.Header.Get("New-Api-User")
+	if apiUserIdStr == "" {
+		c.JSON(http.StatusUnauthorized, gin.H{
+			"success": false,
+			"message": "无权进行此操作，未提供 New-Api-User",
+		})
+		c.Abort()
+		return
+	}
+	apiUserId, err := strconv.Atoi(apiUserIdStr)
+	if err != nil {
+		c.JSON(http.StatusUnauthorized, gin.H{
+			"success": false,
+			"message": "无权进行此操作，New-Api-User 格式错误",
+		})
+		c.Abort()
+		return
 
-		}
-		if id != apiUserId {
-			c.JSON(http.StatusUnauthorized, gin.H{
-				"success": false,
-				"message": "无权进行此操作，与登录用户不匹配，请重新登录",
-			})
-			c.Abort()
-			return
-		}
+	}
+	if id != apiUserId {
+		c.JSON(http.StatusUnauthorized, gin.H{
+			"success": false,
+			"message": "无权进行此操作，New-Api-User 与登录用户不匹配",
+		})
+		c.Abort()
+		return
 	}
 	if status.(int) == common.UserStatusDisabled {
 		c.JSON(http.StatusOK, gin.H{
@@ -201,7 +199,7 @@ func TokenAuth() func(c *gin.Context) {
 			abortWithOpenAiMessage(c, http.StatusUnauthorized, err.Error())
 			return
 		}
-		userEnabled, err := model.CacheIsUserEnabled(token.UserId)
+		userEnabled, err := model.IsUserEnabled(token.UserId, false)
 		if err != nil {
 			abortWithOpenAiMessage(c, http.StatusInternalServerError, err.Error())
 			return

middleware/distributor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/model"
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
+	"one-api/setting"
 	"strconv"
 	"strings"
 	"time"
@@ -39,16 +40,16 @@ func Distribute() func(c *gin.Context) {
 			abortWithOpenAiMessage(c, http.StatusBadRequest, "Invalid request, "+err.Error())
 			return
 		}
-		userGroup, _ := model.CacheGetUserGroup(userId)
+		userGroup, _ := model.GetUserGroup(userId, false)
 		tokenGroup := c.GetString("token_group")
 		if tokenGroup != "" {
 			// check common.UserUsableGroups[userGroup]
-			if _, ok := common.GetUserUsableGroups(userGroup)[tokenGroup]; !ok {
+			if _, ok := setting.GetUserUsableGroups(userGroup)[tokenGroup]; !ok {
 				abortWithOpenAiMessage(c, http.StatusForbidden, fmt.Sprintf("令牌分组 %s 已被禁用", tokenGroup))
 				return
 			}
 			// check group in common.GroupRatio
-			if _, ok := common.GroupRatio[tokenGroup]; !ok {
+			if !setting.ContainsGroupRatio(tokenGroup) {
 				abortWithOpenAiMessage(c, http.StatusForbidden, fmt.Sprintf("分组 %s 已被弃用", tokenGroup))
 				return
 			}

model/ability.go

@@ -3,15 +3,16 @@ package model
 import (
 	"errors"
 	"fmt"
-	"github.com/samber/lo"
-	"gorm.io/gorm"
 	"one-api/common"
 	"strings"
+
+	"github.com/samber/lo"
+	"gorm.io/gorm"
 )
 
 type Ability struct {
 	Group     string  `json:"group" gorm:"type:varchar(64);primaryKey;autoIncrement:false"`
-	Model     string  `json:"model" gorm:"type:varchar(64);primaryKey;autoIncrement:false"`
+	Model     string  `json:"model" gorm:"type:varchar(255);primaryKey;autoIncrement:false"`
 	ChannelId int     `json:"channel_id" gorm:"primaryKey;autoIncrement:false;index"`
 	Enabled   bool    `json:"enabled"`
 	Priority  *int64  `json:"priority" gorm:"bigint;default:0;index"`
@@ -22,10 +23,6 @@ type Ability struct {
 func GetGroupModels(group string) []string {
 	var models []string
 	// Find distinct models
-	groupCol := "`group`"
-	if common.UsingPostgreSQL {
-		groupCol = `"group"`
-	}
 	DB.Table("abilities").Where(groupCol+" = ? and enabled = ?", group, true).Distinct("model").Pluck("model", &models)
 	return models
 }
@@ -44,10 +41,8 @@ func GetAllEnableAbilities() []Ability {
 }
 
 func getPriority(group string, model string, retry int) (int, error) {
-	groupCol := "`group`"
 	trueVal := "1"
 	if common.UsingPostgreSQL {
-		groupCol = `"group"`
 		trueVal = "true"
 	}
 
@@ -80,10 +75,8 @@ func getPriority(group string, model string, retry int) (int, error) {
 }
 
 func getChannelQuery(group string, model string, retry int) *gorm.DB {
-	groupCol := "`group`"
 	trueVal := "1"
 	if common.UsingPostgreSQL {
-		groupCol = `"group"`
 		trueVal = "true"
 	}
 	maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model)
@@ -173,18 +166,67 @@ func (channel *Channel) DeleteAbilities() error {
 
 // UpdateAbilities updates abilities of this channel.
 // Make sure the channel is completed before calling this function.
-func (channel *Channel) UpdateAbilities() error {
-	// A quick and dirty way to update abilities
+func (channel *Channel) UpdateAbilities(tx *gorm.DB) error {
+	isNewTx := false
+	// 如果没有传入事务，创建新的事务
+	if tx == nil {
+		tx = DB.Begin()
+		if tx.Error != nil {
+			return tx.Error
+		}
+		isNewTx = true
+		defer func() {
+			if r := recover(); r != nil {
+				tx.Rollback()
+			}
+		}()
+	}
+
 	// First delete all abilities of this channel
-	err := channel.DeleteAbilities()
+	err := tx.Where("channel_id = ?", channel.Id).Delete(&Ability{}).Error
 	if err != nil {
+		if isNewTx {
+			tx.Rollback()
+		}
 		return err
 	}
+
 	// Then add new abilities
-	err = channel.AddAbilities()
-	if err != nil {
-		return err
+	models_ := strings.Split(channel.Models, ",")
+	groups_ := strings.Split(channel.Group, ",")
+	abilities := make([]Ability, 0, len(models_))
+	for _, model := range models_ {
+		for _, group := range groups_ {
+			ability := Ability{
+				Group:     group,
+				Model:     model,
+				ChannelId: channel.Id,
+				Enabled:   channel.Status == common.ChannelStatusEnabled,
+				Priority:  channel.Priority,
+				Weight:    uint(channel.GetWeight()),
+				Tag:       channel.Tag,
+			}
+			abilities = append(abilities, ability)
+		}
 	}
+
+	if len(abilities) > 0 {
+		for _, chunk := range lo.Chunk(abilities, 50) {
+			err = tx.Create(&chunk).Error
+			if err != nil {
+				if isNewTx {
+					tx.Rollback()
+				}
+				return err
+			}
+		}
+	}
+
+	// 如果是新创建的事务，需要提交
+	if isNewTx {
+		return tx.Commit().Error
+	}
+
 	return nil
 }
 
@@ -236,7 +278,6 @@ func FixAbility() (int, error) {
 		return 0, err
 	}
 	var channels []Channel
-
 	if len(abilityChannelIds) == 0 {
 		err = DB.Find(&channels).Error
 	} else {
@@ -246,7 +287,7 @@ func FixAbility() (int, error) {
 		return 0, err
 	}
 	for _, channel := range channels {
-		err := channel.UpdateAbilities()
+		err := channel.UpdateAbilities(nil)
 		if err != nil {
 			common.SysError(fmt.Sprintf("Update abilities of channel %d failed: %s", channel.Id, err.Error()))
 		} else {

model/cache.go

@@ -1,209 +1,115 @@
 package model
 
 import (
-	"encoding/json"
 	"errors"
 	"fmt"
 	"math/rand"
 	"one-api/common"
 	"sort"
-	"strconv"
 	"strings"
 	"sync"
 	"time"
 )
 
-var (
-	TokenCacheSeconds         = common.SyncFrequency
-	UserId2GroupCacheSeconds  = common.SyncFrequency
-	UserId2QuotaCacheSeconds  = common.SyncFrequency
-	UserId2StatusCacheSeconds = common.SyncFrequency
-)
-
-// 仅用于定时同步缓存
-var token2UserId = make(map[string]int)
-var token2UserIdLock sync.RWMutex
-
-func cacheSetToken(token *Token) error {
-	jsonBytes, err := json.Marshal(token)
-	if err != nil {
-		return err
-	}
-	err = common.RedisSet(fmt.Sprintf("token:%s", token.Key), string(jsonBytes), time.Duration(TokenCacheSeconds)*time.Second)
-	if err != nil {
-		common.SysError(fmt.Sprintf("failed to set token %s to redis: %s", token.Key, err.Error()))
-		return err
-	}
-	token2UserIdLock.Lock()
-	defer token2UserIdLock.Unlock()
-	token2UserId[token.Key] = token.UserId
-	return nil
-}
-
-// CacheGetTokenByKey 从缓存中获取 token 并续期时间，如果缓存中不存在，则从数据库中获取
-func CacheGetTokenByKey(key string) (*Token, error) {
-	if !common.RedisEnabled {
-		return GetTokenByKey(key)
-	}
-	var token *Token
-	tokenObjectString, err := common.RedisGet(fmt.Sprintf("token:%s", key))
-	if err != nil {
-		// 如果缓存中不存在，则从数据库中获取
-		token, err = GetTokenByKey(key)
-		if err != nil {
-			return nil, err
-		}
-		err = cacheSetToken(token)
-		return token, nil
-	}
-	// 如果缓存中存在，则续期时间
-	err = common.RedisExpire(fmt.Sprintf("token:%s", key), time.Duration(TokenCacheSeconds)*time.Second)
-	err = json.Unmarshal([]byte(tokenObjectString), &token)
-	return token, err
-}
-
-func SyncTokenCache(frequency int) {
-	for {
-		time.Sleep(time.Duration(frequency) * time.Second)
-		common.SysLog("syncing tokens from database")
-		token2UserIdLock.Lock()
-		// 从token2UserId中获取所有的key
-		var copyToken2UserId = make(map[string]int)
-		for s, i := range token2UserId {
-			copyToken2UserId[s] = i
-		}
-		token2UserId = make(map[string]int)
-		token2UserIdLock.Unlock()
-
-		for key := range copyToken2UserId {
-			token, err := GetTokenByKey(key)
-			if err != nil {
-				// 如果数据库中不存在，则删除缓存
-				common.SysError(fmt.Sprintf("failed to get token %s from database: %s", key, err.Error()))
-				//delete redis
-				err := common.RedisDel(fmt.Sprintf("token:%s", key))
-				if err != nil {
-					common.SysError(fmt.Sprintf("failed to delete token %s from redis: %s", key, err.Error()))
-				}
-			} else {
-				// 如果数据库中存在，先检查redis
-				_, err = common.RedisGet(fmt.Sprintf("token:%s", key))
-				if err != nil {
-					// 如果redis中不存在，则跳过
-					continue
-				}
-				err = cacheSetToken(token)
-				if err != nil {
-					common.SysError(fmt.Sprintf("failed to update token %s to redis: %s", key, err.Error()))
-				}
-			}
-		}
-	}
-}
-
-func CacheGetUserGroup(id int) (group string, err error) {
-	if !common.RedisEnabled {
-		return GetUserGroup(id)
-	}
-	group, err = common.RedisGet(fmt.Sprintf("user_group:%d", id))
-	if err != nil {
-		group, err = GetUserGroup(id)
-		if err != nil {
-			return "", err
-		}
-		err = common.RedisSet(fmt.Sprintf("user_group:%d", id), group, time.Duration(UserId2GroupCacheSeconds)*time.Second)
-		if err != nil {
-			common.SysError("Redis set user group error: " + err.Error())
-		}
-	}
-	return group, err
-}
-
-func CacheGetUsername(id int) (username string, err error) {
-	if !common.RedisEnabled {
-		return GetUsernameById(id)
-	}
-	username, err = common.RedisGet(fmt.Sprintf("user_name:%d", id))
-	if err != nil {
-		username, err = GetUsernameById(id)
-		if err != nil {
-			return "", err
-		}
-		err = common.RedisSet(fmt.Sprintf("user_name:%d", id), username, time.Duration(UserId2GroupCacheSeconds)*time.Second)
-		if err != nil {
-			common.SysError("Redis set user group error: " + err.Error())
-		}
-	}
-	return username, err
-}
-
-func CacheGetUserQuota(id int) (quota int, err error) {
-	if !common.RedisEnabled {
-		return GetUserQuota(id)
-	}
-	quotaString, err := common.RedisGet(fmt.Sprintf("user_quota:%d", id))
-	if err != nil {
-		quota, err = GetUserQuota(id)
-		if err != nil {
-			return 0, err
-		}
-		err = common.RedisSet(fmt.Sprintf("user_quota:%d", id), fmt.Sprintf("%d", quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
-		if err != nil {
-			common.SysError("Redis set user quota error: " + err.Error())
-		}
-		return quota, err
-	}
-	quota, err = strconv.Atoi(quotaString)
-	return quota, err
-}
-
-func CacheUpdateUserQuota(id int) error {
-	if !common.RedisEnabled {
-		return nil
-	}
-	quota, err := GetUserQuota(id)
-	if err != nil {
-		return err
-	}
-	return cacheSetUserQuota(id, quota)
-}
-
-func cacheSetUserQuota(id int, quota int) error {
-	err := common.RedisSet(fmt.Sprintf("user_quota:%d", id), fmt.Sprintf("%d", quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
-	return err
-}
-
-func CacheDecreaseUserQuota(id int, quota int) error {
-	if !common.RedisEnabled {
-		return nil
-	}
-	err := common.RedisDecrease(fmt.Sprintf("user_quota:%d", id), int64(quota))
-	return err
-}
-
-func CacheIsUserEnabled(userId int) (bool, error) {
-	if !common.RedisEnabled {
-		return IsUserEnabled(userId)
-	}
-	enabled, err := common.RedisGet(fmt.Sprintf("user_enabled:%d", userId))
-	if err == nil {
-		return enabled == "1", nil
-	}
-
-	userEnabled, err := IsUserEnabled(userId)
-	if err != nil {
-		return false, err
-	}
-	enabled = "0"
-	if userEnabled {
-		enabled = "1"
-	}
-	err = common.RedisSet(fmt.Sprintf("user_enabled:%d", userId), enabled, time.Duration(UserId2StatusCacheSeconds)*time.Second)
-	if err != nil {
-		common.SysError("Redis set user enabled error: " + err.Error())
-	}
-	return userEnabled, err
-}
+//func CacheGetUserGroup(id int) (group string, err error) {
+//	if !common.RedisEnabled {
+//		return GetUserGroup(id)
+//	}
+//	group, err = common.RedisGet(fmt.Sprintf("user_group:%d", id))
+//	if err != nil {
+//		group, err = GetUserGroup(id)
+//		if err != nil {
+//			return "", err
+//		}
+//		err = common.RedisSet(fmt.Sprintf("user_group:%d", id), group, time.Duration(constant.UserId2GroupCacheSeconds)*time.Second)
+//		if err != nil {
+//			common.SysError("Redis set user group error: " + err.Error())
+//		}
+//	}
+//	return group, err
+//}
+//
+//func CacheGetUsername(id int) (username string, err error) {
+//	if !common.RedisEnabled {
+//		return GetUsernameById(id)
+//	}
+//	username, err = common.RedisGet(fmt.Sprintf("user_name:%d", id))
+//	if err != nil {
+//		username, err = GetUsernameById(id)
+//		if err != nil {
+//			return "", err
+//		}
+//		err = common.RedisSet(fmt.Sprintf("user_name:%d", id), username, time.Duration(constant.UserId2GroupCacheSeconds)*time.Second)
+//		if err != nil {
+//			common.SysError("Redis set user group error: " + err.Error())
+//		}
+//	}
+//	return username, err
+//}
+//
+//func CacheGetUserQuota(id int) (quota int, err error) {
+//	if !common.RedisEnabled {
+//		return GetUserQuota(id)
+//	}
+//	quotaString, err := common.RedisGet(fmt.Sprintf("user_quota:%d", id))
+//	if err != nil {
+//		quota, err = GetUserQuota(id)
+//		if err != nil {
+//			return 0, err
+//		}
+//		return quota, nil
+//	}
+//	quota, err = strconv.Atoi(quotaString)
+//	return quota, nil
+//}
+//
+//func CacheUpdateUserQuota(id int) error {
+//	if !common.RedisEnabled {
+//		return nil
+//	}
+//	quota, err := GetUserQuota(id)
+//	if err != nil {
+//		return err
+//	}
+//	return cacheSetUserQuota(id, quota)
+//}
+//
+//func cacheSetUserQuota(id int, quota int) error {
+//	err := common.RedisSet(fmt.Sprintf("user_quota:%d", id), fmt.Sprintf("%d", quota), time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second)
+//	return err
+//}
+//
+//func CacheDecreaseUserQuota(id int, quota int) error {
+//	if !common.RedisEnabled {
+//		return nil
+//	}
+//	err := common.RedisDecrease(fmt.Sprintf("user_quota:%d", id), int64(quota))
+//	return err
+//}
+//
+//func CacheIsUserEnabled(userId int) (bool, error) {
+//	if !common.RedisEnabled {
+//		return IsUserEnabled(userId)
+//	}
+//	enabled, err := common.RedisGet(fmt.Sprintf("user_enabled:%d", userId))
+//	if err == nil {
+//		return enabled == "1", nil
+//	}
+//
+//	userEnabled, err := IsUserEnabled(userId)
+//	if err != nil {
+//		return false, err
+//	}
+//	enabled = "0"
+//	if userEnabled {
+//		enabled = "1"
+//	}
+//	err = common.RedisSet(fmt.Sprintf("user_enabled:%d", userId), enabled, time.Duration(constant.UserId2StatusCacheSeconds)*time.Second)
+//	if err != nil {
+//		common.SysError("Redis set user enabled error: " + err.Error())
+//	}
+//	return userEnabled, err
+//}
 
 var group2model2channels map[string]map[string][]*Channel
 var channelsIDM map[int]*Channel
@@ -344,12 +250,12 @@ func CacheGetChannel(id int) (*Channel, error) {
 }
 
 func CacheUpdateChannelStatus(id int, status int) {
-    if (!common.MemoryCacheEnabled) {
-        return
-    }
-    channelSyncLock.Lock()
-    defer channelSyncLock.Unlock()
-    if channel, ok := channelsIDM[id]; ok {
-        channel.Status = status
-    }
+	if !common.MemoryCacheEnabled {
+		return
+	}
+	channelSyncLock.Lock()
+	defer channelSyncLock.Unlock()
+	if channel, ok := channelsIDM[id]; ok {
+		channel.Status = status
+	}
 }

model/channel.go

@@ -28,7 +28,7 @@ type Channel struct {
 	Models             string  `json:"models"`
 	Group              string  `json:"group" gorm:"type:varchar(64);default:'default'"`
 	UsedQuota          int64   `json:"used_quota" gorm:"bigint;default:0"`
-	ModelMapping       *string `json:"model_mapping" gorm:"type:varchar(1024);default:''"`
+	ModelMapping       *string `json:"model_mapping" gorm:"type:text"`
 	//MaxInputTokens     *int    `json:"max_input_tokens" gorm:"default:0"`
 	StatusCodeMapping *string `json:"status_code_mapping" gorm:"type:varchar(1024);default:''"`
 	Priority          *int64  `json:"priority" gorm:"bigint;default:0"`
@@ -114,14 +114,11 @@ func GetChannelsByTag(tag string, idSort bool) ([]*Channel, error) {
 
 func SearchChannels(keyword string, group string, model string, idSort bool) ([]*Channel, error) {
 	var channels []*Channel
-	keyCol := "`key`"
-	groupCol := "`group`"
 	modelsCol := "`models`"
 
 	// 如果是 PostgreSQL，使用双引号
 	if common.UsingPostgreSQL {
 		keyCol = `"key"`
-		groupCol = `"group"`
 		modelsCol = `"models"`
 	}
 
@@ -257,7 +254,7 @@ func (channel *Channel) Update() error {
 		return err
 	}
 	DB.Model(channel).First(channel, "id = ?", channel.Id)
-	err = channel.UpdateAbilities()
+	err = channel.UpdateAbilities(nil)
 	return err
 }
 
@@ -389,7 +386,7 @@ func EditChannelByTag(tag string, newTag *string, modelMapping *string, models *
 		channels, err := GetChannelsByTag(updatedTag, false)
 		if err == nil {
 			for _, channel := range channels {
-				err = channel.UpdateAbilities()
+				err = channel.UpdateAbilities(nil)
 				if err != nil {
 					common.SysError("failed to update abilities: " + err.Error())
 				}
@@ -437,14 +434,10 @@ func GetPaginatedTags(offset int, limit int) ([]*string, error) {
 
 func SearchTags(keyword string, group string, model string, idSort bool) ([]*string, error) {
 	var tags []*string
-	keyCol := "`key`"
-	groupCol := "`group`"
 	modelsCol := "`models`"
 
 	// 如果是 PostgreSQL，使用双引号
 	if common.UsingPostgreSQL {
-		keyCol = `"key"`
-		groupCol = `"group"`
 		modelsCol = `"models"`
 	}
 
@@ -509,3 +502,42 @@ func (channel *Channel) SetSetting(setting map[string]interface{}) {
 	}
 	channel.Setting = string(settingBytes)
 }
+
+func GetChannelsByIds(ids []int) ([]*Channel, error) {
+	var channels []*Channel
+	err := DB.Where("id in (?)", ids).Find(&channels).Error
+	return channels, err
+}
+
+func BatchSetChannelTag(ids []int, tag *string) error {
+	// 开启事务
+	tx := DB.Begin()
+	if tx.Error != nil {
+		return tx.Error
+	}
+
+	// 更新标签
+	err := tx.Model(&Channel{}).Where("id in (?)", ids).Update("tag", tag).Error
+	if err != nil {
+		tx.Rollback()
+		return err
+	}
+
+	// update ability status
+	channels, err := GetChannelsByIds(ids)
+	if err != nil {
+		tx.Rollback()
+		return err
+	}
+
+	for _, channel := range channels {
+		err = channel.UpdateAbilities(tx)
+		if err != nil {
+			tx.Rollback()
+			return err
+		}
+	}
+
+	// 提交事务
+	return tx.Commit().Error
+}

model/log.go

@@ -12,16 +12,6 @@ import (
 	"gorm.io/gorm"
 )
 
-var groupCol string
-
-func init() {
-	if common.UsingPostgreSQL {
-		groupCol = `"group"`
-	} else {
-		groupCol = "`group`"
-	}
-}
-
 type Log struct {
 	Id               int    `json:"id" gorm:"index:idx_created_at_id,priority:1"`
 	UserId           int    `json:"user_id" gorm:"index"`
@@ -37,6 +27,7 @@ type Log struct {
 	UseTime          int    `json:"use_time" gorm:"default:0"`
 	IsStream         bool   `json:"is_stream" gorm:"default:false"`
 	ChannelId        int    `json:"channel" gorm:"index"`
+	ChannelName      string `json:"channel_name" gorm:"->"`
 	TokenId          int    `json:"token_id" gorm:"default:0;index"`
 	Group            string `json:"group" gorm:"index"`
 	Other            string `json:"other"`
@@ -50,16 +41,31 @@ const (
 	LogTypeSystem
 )
 
+func formatUserLogs(logs []*Log) {
+	for i := range logs {
+		logs[i].ChannelName = ""
+		var otherMap map[string]interface{}
+		otherMap = common.StrToMap(logs[i].Other)
+		if otherMap != nil {
+			// delete admin
+			delete(otherMap, "admin_info")
+		}
+		logs[i].Other = common.MapToJsonStr(otherMap)
+		logs[i].Id = logs[i].Id % 1024
+	}
+}
+
 func GetLogByKey(key string) (logs []*Log, err error) {
 	if os.Getenv("LOG_SQL_DSN") != "" {
 		var tk Token
-		if err = DB.Model(&Token{}).Where("`key`=?", strings.TrimPrefix(key, "sk-")).First(&tk).Error; err != nil {
+		if err = DB.Model(&Token{}).Where(keyCol+"=?", strings.TrimPrefix(key, "sk-")).First(&tk).Error; err != nil {
 			return nil, err
 		}
 		err = LOG_DB.Model(&Log{}).Where("token_id=?", tk.Id).Find(&logs).Error
 	} else {
 		err = LOG_DB.Joins("left join tokens on tokens.id = logs.token_id").Where("tokens.key = ?", strings.TrimPrefix(key, "sk-")).Find(&logs).Error
 	}
+	formatUserLogs(logs)
 	return logs, err
 }
 
@@ -67,7 +73,7 @@ func RecordLog(userId int, logType int, content string) {
 	if logType == LogTypeConsume && !common.LogConsumeEnabled {
 		return
 	}
-	username, _ := CacheGetUsername(userId)
+	username, _ := GetUsernameById(userId, false)
 	log := &Log{
 		UserId:    userId,
 		Username:  username,
@@ -88,7 +94,7 @@ func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptToke
 	if !common.LogConsumeEnabled {
 		return
 	}
-	username, _ := CacheGetUsername(userId)
+	username, _ := GetUsernameById(userId, false)
 	otherStr := common.MapToJsonStr(other)
 	log := &Log{
 		UserId:           userId,
@@ -124,34 +130,38 @@ func GetAllLogs(logType int, startTimestamp int64, endTimestamp int64, modelName
 	if logType == LogTypeUnknown {
 		tx = LOG_DB
 	} else {
-		tx = LOG_DB.Where("type = ?", logType)
+		tx = LOG_DB.Where("logs.type = ?", logType)
 	}
+
+	tx = tx.Joins("LEFT JOIN channels ON logs.channel_id = channels.id")
+	tx = tx.Select("logs.*, channels.name as channel_name")
+
 	if modelName != "" {
-		tx = tx.Where("model_name like ?", modelName)
+		tx = tx.Where("logs.model_name like ?", modelName)
 	}
 	if username != "" {
-		tx = tx.Where("username = ?", username)
+		tx = tx.Where("logs.username = ?", username)
 	}
 	if tokenName != "" {
-		tx = tx.Where("token_name = ?", tokenName)
+		tx = tx.Where("logs.token_name = ?", tokenName)
 	}
 	if startTimestamp != 0 {
-		tx = tx.Where("created_at >= ?", startTimestamp)
+		tx = tx.Where("logs.created_at >= ?", startTimestamp)
 	}
 	if endTimestamp != 0 {
-		tx = tx.Where("created_at <= ?", endTimestamp)
+		tx = tx.Where("logs.created_at <= ?", endTimestamp)
 	}
 	if channel != 0 {
-		tx = tx.Where("channel_id = ?", channel)
+		tx = tx.Where("logs.channel_id = ?", channel)
 	}
 	if group != "" {
-		tx = tx.Where(groupCol+" = ?", group)
+		tx = tx.Where("logs."+groupCol+" = ?", group)
 	}
 	err = tx.Model(&Log{}).Count(&total).Error
 	if err != nil {
 		return nil, 0, err
 	}
-	err = tx.Order("id desc").Limit(num).Offset(startIdx).Find(&logs).Error
+	err = tx.Order("logs.id desc").Limit(num).Offset(startIdx).Find(&logs).Error
 	if err != nil {
 		return nil, 0, err
 	}
@@ -161,39 +171,35 @@ func GetAllLogs(logType int, startTimestamp int64, endTimestamp int64, modelName
 func GetUserLogs(userId int, logType int, startTimestamp int64, endTimestamp int64, modelName string, tokenName string, startIdx int, num int, group string) (logs []*Log, total int64, err error) {
 	var tx *gorm.DB
 	if logType == LogTypeUnknown {
-		tx = LOG_DB.Where("user_id = ?", userId)
+		tx = LOG_DB.Where("logs.user_id = ?", userId)
 	} else {
-		tx = LOG_DB.Where("user_id = ? and type = ?", userId, logType)
+		tx = LOG_DB.Where("logs.user_id = ? and logs.type = ?", userId, logType)
 	}
+
+	tx = tx.Joins("LEFT JOIN channels ON logs.channel_id = channels.id")
+	tx = tx.Select("logs.*, channels.name as channel_name")
+
 	if modelName != "" {
-		tx = tx.Where("model_name like ?", modelName)
+		tx = tx.Where("logs.model_name like ?", modelName)
 	}
 	if tokenName != "" {
-		tx = tx.Where("token_name = ?", tokenName)
+		tx = tx.Where("logs.token_name = ?", tokenName)
 	}
 	if startTimestamp != 0 {
-		tx = tx.Where("created_at >= ?", startTimestamp)
+		tx = tx.Where("logs.created_at >= ?", startTimestamp)
 	}
 	if endTimestamp != 0 {
-		tx = tx.Where("created_at <= ?", endTimestamp)
+		tx = tx.Where("logs.created_at <= ?", endTimestamp)
 	}
 	if group != "" {
-		tx = tx.Where(groupCol+" = ?", group)
+		tx = tx.Where("logs."+groupCol+" = ?", group)
 	}
 	err = tx.Model(&Log{}).Count(&total).Error
 	if err != nil {
 		return nil, 0, err
 	}
-	err = tx.Order("id desc").Limit(num).Offset(startIdx).Omit("id").Find(&logs).Error
-	for i := range logs {
-		var otherMap map[string]interface{}
-		otherMap = common.StrToMap(logs[i].Other)
-		if otherMap != nil {
-			// delete admin
-			delete(otherMap, "admin_info")
-		}
-		logs[i].Other = common.MapToJsonStr(otherMap)
-	}
+	err = tx.Order("logs.id desc").Limit(num).Offset(startIdx).Find(&logs).Error
+	formatUserLogs(logs)
 	return logs, total, err
 }
 
@@ -203,7 +209,8 @@ func SearchAllLogs(keyword string) (logs []*Log, err error) {
 }
 
 func SearchUserLogs(userId int, keyword string) (logs []*Log, err error) {
-	err = LOG_DB.Where("user_id = ? and type = ?", userId, keyword).Order("id desc").Limit(common.MaxRecentItems).Omit("id").Find(&logs).Error
+	err = LOG_DB.Where("user_id = ? and type = ?", userId, keyword).Order("id desc").Limit(common.MaxRecentItems).Find(&logs).Error
+	formatUserLogs(logs)
 	return logs, err
 }
 

model/main.go

@@ -13,6 +13,20 @@ import (
 	"time"
 )
 
+var groupCol string
+var keyCol string
+
+func initCol() {
+	if common.UsingPostgreSQL {
+		groupCol = `"group"`
+		keyCol = `"key"`
+
+	} else {
+		groupCol = "`group`"
+		keyCol = "`key`"
+	}
+}
+
 var DB *gorm.DB
 
 var LOG_DB *gorm.DB
@@ -41,6 +55,9 @@ func createRootAccountIfNeed() error {
 }
 
 func chooseDB(envName string) (*gorm.DB, error) {
+	defer func() {
+		initCol()
+	}()
 	dsn := os.Getenv(envName)
 	if dsn != "" {
 		if strings.HasPrefix(dsn, "postgres://") {
@@ -102,12 +119,9 @@ func InitDB() (err error) {
 		if !common.IsMasterNode {
 			return nil
 		}
-		//if common.UsingMySQL {
-		//	_, _ = sqlDB.Exec("DROP INDEX idx_channels_key ON channels;")             // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY action VARCHAR(40);")   // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY progress VARCHAR(30);") // TODO: delete this line when most users have upgraded
-		//	_, _ = sqlDB.Exec("ALTER TABLE midjourneys MODIFY status VARCHAR(20);")   // TODO: delete this line when most users have upgraded
-		//}
+		if common.UsingMySQL {
+			_, _ = sqlDB.Exec("ALTER TABLE channels MODIFY model_mapping TEXT;") // TODO: delete this line when most users have upgraded
+		}
 		common.SysLog("database migration started")
 		err = migrateDB()
 		return err

model/option.go

@@ -87,8 +87,8 @@ func InitOptionMap() {
 	common.OptionMap["PreConsumedQuota"] = strconv.Itoa(common.PreConsumedQuota)
 	common.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
 	common.OptionMap["ModelPrice"] = common.ModelPrice2JSONString()
-	common.OptionMap["GroupRatio"] = common.GroupRatio2JSONString()
-	common.OptionMap["UserUsableGroups"] = common.UserUsableGroups2JSONString()
+	common.OptionMap["GroupRatio"] = setting.GroupRatio2JSONString()
+	common.OptionMap["UserUsableGroups"] = setting.UserUsableGroups2JSONString()
 	common.OptionMap["CompletionRatio"] = common.CompletionRatio2JSONString()
 	common.OptionMap["TopUpLink"] = common.TopUpLink
 	common.OptionMap["ChatLink"] = common.ChatLink
@@ -104,6 +104,7 @@ func InitOptionMap() {
 	common.OptionMap["MjForwardUrlEnabled"] = strconv.FormatBool(setting.MjForwardUrlEnabled)
 	common.OptionMap["MjActionCheckSuccessEnabled"] = strconv.FormatBool(setting.MjActionCheckSuccessEnabled)
 	common.OptionMap["CheckSensitiveEnabled"] = strconv.FormatBool(setting.CheckSensitiveEnabled)
+	common.OptionMap["DemoSiteEnabled"] = strconv.FormatBool(setting.DemoSiteEnabled)
 	common.OptionMap["CheckSensitiveOnPromptEnabled"] = strconv.FormatBool(setting.CheckSensitiveOnPromptEnabled)
 	//common.OptionMap["CheckSensitiveOnCompletionEnabled"] = strconv.FormatBool(constant.CheckSensitiveOnCompletionEnabled)
 	common.OptionMap["StopOnSensitiveEnabled"] = strconv.FormatBool(setting.StopOnSensitiveEnabled)
@@ -220,6 +221,8 @@ func updateOptionMap(key string, value string) (err error) {
 			setting.MjActionCheckSuccessEnabled = boolValue
 		case "CheckSensitiveEnabled":
 			setting.CheckSensitiveEnabled = boolValue
+		case "DemoSiteEnabled":
+			setting.DemoSiteEnabled = boolValue
 		case "CheckSensitiveOnPromptEnabled":
 			setting.CheckSensitiveOnPromptEnabled = boolValue
 		//case "CheckSensitiveOnCompletionEnabled":
@@ -313,9 +316,9 @@ func updateOptionMap(key string, value string) (err error) {
 	case "ModelRatio":
 		err = common.UpdateModelRatioByJSONString(value)
 	case "GroupRatio":
-		err = common.UpdateGroupRatioByJSONString(value)
+		err = setting.UpdateGroupRatioByJSONString(value)
 	case "UserUsableGroups":
-		err = common.UpdateUserUsableGroupsByJSONString(value)
+		err = setting.UpdateUserUsableGroupsByJSONString(value)
 	case "CompletionRatio":
 		err = common.UpdateCompletionRatioByJSONString(value)
 	case "ModelPrice":

model/redemption.go

@@ -3,8 +3,10 @@ package model
 import (
 	"errors"
 	"fmt"
-	"gorm.io/gorm"
 	"one-api/common"
+	"strconv"
+
+	"gorm.io/gorm"
 )
 
 type Redemption struct {
@@ -21,16 +23,80 @@ type Redemption struct {
 	DeletedAt    gorm.DeletedAt `gorm:"index"`
 }
 
-func GetAllRedemptions(startIdx int, num int) ([]*Redemption, error) {
-	var redemptions []*Redemption
-	var err error
-	err = DB.Order("id desc").Limit(num).Offset(startIdx).Find(&redemptions).Error
-	return redemptions, err
+func GetAllRedemptions(startIdx int, num int) (redemptions []*Redemption, total int64, err error) {
+	// 开始事务
+	tx := DB.Begin()
+	if tx.Error != nil {
+		return nil, 0, tx.Error
+	}
+	defer func() {
+		if r := recover(); r != nil {
+			tx.Rollback()
+		}
+	}()
+
+	// 获取总数
+	err = tx.Model(&Redemption{}).Count(&total).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// 获取分页数据
+	err = tx.Order("id desc").Limit(num).Offset(startIdx).Find(&redemptions).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// 提交事务
+	if err = tx.Commit().Error; err != nil {
+		return nil, 0, err
+	}
+
+	return redemptions, total, nil
 }
 
-func SearchRedemptions(keyword string) (redemptions []*Redemption, err error) {
-	err = DB.Where("id = ? or name LIKE ?", keyword, keyword+"%").Find(&redemptions).Error
-	return redemptions, err
+func SearchRedemptions(keyword string, startIdx int, num int) (redemptions []*Redemption, total int64, err error) {
+	tx := DB.Begin()
+	if tx.Error != nil {
+		return nil, 0, tx.Error
+	}
+	defer func() {
+		if r := recover(); r != nil {
+			tx.Rollback()
+		}
+	}()
+
+	// Build query based on keyword type
+	query := tx.Model(&Redemption{})
+
+	// Only try to convert to ID if the string represents a valid integer
+	if id, err := strconv.Atoi(keyword); err == nil {
+		query = query.Where("id = ? OR name LIKE ?", id, keyword+"%")
+	} else {
+		query = query.Where("name LIKE ?", keyword+"%")
+	}
+
+	// Get total count
+	err = query.Count(&total).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// Get paginated data
+	err = query.Order("id desc").Limit(num).Offset(startIdx).Find(&redemptions).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	if err = tx.Commit().Error; err != nil {
+		return nil, 0, err
+	}
+
+	return redemptions, total, nil
 }
 
 func GetRedemptionById(id int) (*Redemption, error) {

model/token.go

@@ -3,6 +3,7 @@ package model
 import (
 	"errors"
 	"fmt"
+	"github.com/bytedance/gopkg/util/gopool"
 	"gorm.io/gorm"
 	"one-api/common"
 	relaycommon "one-api/relay/common"
@@ -30,6 +31,10 @@ type Token struct {
 	DeletedAt          gorm.DeletedAt `gorm:"index"`
 }
 
+func (token *Token) Clean() {
+	token.Key = ""
+}
+
 func (token *Token) GetIpLimitsMap() map[string]any {
 	// delete empty spaces
 	//split with \n
@@ -63,7 +68,7 @@ func SearchUserTokens(userId int, keyword string, token string) (tokens []*Token
 	if token != "" {
 		token = strings.Trim(token, "sk-")
 	}
-	err = DB.Where("user_id = ?", userId).Where("name LIKE ?", "%"+keyword+"%").Where("`key` LIKE ?", "%"+token+"%").Find(&tokens).Error
+	err = DB.Where("user_id = ?", userId).Where("name LIKE ?", "%"+keyword+"%").Where(keyCol+" LIKE ?", "%"+token+"%").Find(&tokens).Error
 	return tokens, err
 }
 
@@ -71,7 +76,7 @@ func ValidateUserToken(key string) (token *Token, err error) {
 	if key == "" {
 		return nil, errors.New("未提供令牌")
 	}
-	token, err = CacheGetTokenByKey(key)
+	token, err = GetTokenByKey(key, false)
 	if err == nil {
 		if token.Status == common.TokenStatusExhausted {
 			keyPrefix := key[:3]
@@ -128,22 +133,38 @@ func GetTokenById(id int) (*Token, error) {
 	token := Token{Id: id}
 	var err error = nil
 	err = DB.First(&token, "id = ?", id).Error
-	if err != nil {
-		if common.RedisEnabled {
-			go cacheSetToken(&token)
-		}
+	if shouldUpdateRedis(true, err) {
+		gopool.Go(func() {
+			if err := cacheSetToken(token); err != nil {
+				common.SysError("failed to update user status cache: " + err.Error())
+			}
+		})
 	}
 	return &token, err
 }
 
-func GetTokenByKey(key string) (*Token, error) {
-	keyCol := "`key`"
-	if common.UsingPostgreSQL {
-		keyCol = `"key"`
+func GetTokenByKey(key string, fromDB bool) (token *Token, err error) {
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) && token != nil {
+			gopool.Go(func() {
+				if err := cacheSetToken(*token); err != nil {
+					common.SysError("failed to update user status cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		// Try Redis first
+		token, err := cacheGetTokenByKey(key)
+		if err == nil {
+			return token, nil
+		}
+		// Don't return error - fall through to DB
 	}
-	var token Token
-	err := DB.Where(keyCol+" = ?", key).First(&token).Error
-	return &token, err
+	fromDB = true
+	err = DB.Where(keyCol+" = ?", key).First(&token).Error
+	return token, err
 }
 
 func (token *Token) Insert() error {
@@ -153,20 +174,48 @@ func (token *Token) Insert() error {
 }
 
 // Update Make sure your token's fields is completed, because this will update non-zero values
-func (token *Token) Update() error {
-	var err error
+func (token *Token) Update() (err error) {
+	defer func() {
+		if shouldUpdateRedis(true, err) {
+			gopool.Go(func() {
+				err := cacheSetToken(*token)
+				if err != nil {
+					common.SysError("failed to update token cache: " + err.Error())
+				}
+			})
+		}
+	}()
 	err = DB.Model(token).Select("name", "status", "expired_time", "remain_quota", "unlimited_quota",
 		"model_limits_enabled", "model_limits", "allow_ips", "group").Updates(token).Error
 	return err
 }
 
-func (token *Token) SelectUpdate() error {
+func (token *Token) SelectUpdate() (err error) {
+	defer func() {
+		if shouldUpdateRedis(true, err) {
+			gopool.Go(func() {
+				err := cacheSetToken(*token)
+				if err != nil {
+					common.SysError("failed to update token cache: " + err.Error())
+				}
+			})
+		}
+	}()
 	// This can update zero values
 	return DB.Model(token).Select("accessed_time", "status").Updates(token).Error
 }
 
-func (token *Token) Delete() error {
-	var err error
+func (token *Token) Delete() (err error) {
+	defer func() {
+		if shouldUpdateRedis(true, err) {
+			gopool.Go(func() {
+				err := cacheDeleteToken(token.Key)
+				if err != nil {
+					common.SysError("failed to delete token cache: " + err.Error())
+				}
+			})
+		}
+	}()
 	err = DB.Delete(token).Error
 	return err
 }
@@ -214,10 +263,18 @@ func DeleteTokenById(id int, userId int) (err error) {
 	return token.Delete()
 }
 
-func IncreaseTokenQuota(id int, quota int) (err error) {
+func IncreaseTokenQuota(id int, key string, quota int) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
+	if common.RedisEnabled {
+		gopool.Go(func() {
+			err := cacheIncrTokenQuota(key, int64(quota))
+			if err != nil {
+				common.SysError("failed to increase token quota: " + err.Error())
+			}
+		})
+	}
 	if common.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeTokenQuota, id, quota)
 		return nil
@@ -236,10 +293,18 @@ func increaseTokenQuota(id int, quota int) (err error) {
 	return err
 }
 
-func DecreaseTokenQuota(id int, quota int) (err error) {
+func DecreaseTokenQuota(id int, key string, quota int) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
+	if common.RedisEnabled {
+		gopool.Go(func() {
+			err := cacheDecrTokenQuota(key, int64(quota))
+			if err != nil {
+				common.SysError("failed to decrease token quota: " + err.Error())
+			}
+		})
+	}
 	if common.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeTokenQuota, id, -quota)
 		return nil
@@ -258,37 +323,31 @@ func decreaseTokenQuota(id int, quota int) (err error) {
 	return err
 }
 
-func PreConsumeTokenQuota(relayInfo *relaycommon.RelayInfo, quota int) (userQuota int, err error) {
+func PreConsumeTokenQuota(relayInfo *relaycommon.RelayInfo, quota int) error {
 	if quota < 0 {
-		return 0, errors.New("quota 不能为负数！")
+		return errors.New("quota 不能为负数！")
 	}
-	if !relayInfo.IsPlayground {
-		token, err := GetTokenById(relayInfo.TokenId)
-		if err != nil {
-			return 0, err
-		}
-		if !token.UnlimitedQuota && token.RemainQuota < quota {
-			return 0, errors.New("令牌额度不足")
-		}
+	if relayInfo.IsPlayground {
+		return nil
 	}
-	userQuota, err = GetUserQuota(relayInfo.UserId)
+	//if relayInfo.TokenUnlimited {
+	//	return nil
+	//}
+	token, err := GetTokenById(relayInfo.TokenId)
 	if err != nil {
-		return 0, err
+		return err
 	}
-	if userQuota < quota {
-		return 0, errors.New(fmt.Sprintf("用户额度不足，剩余额度为 %d", userQuota))
+	if !relayInfo.TokenUnlimited && token.RemainQuota < quota {
+		return errors.New("令牌额度不足")
 	}
-	if !relayInfo.IsPlayground {
-		err = DecreaseTokenQuota(relayInfo.TokenId, quota)
-		if err != nil {
-			return 0, err
-		}
+	err = DecreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, quota)
+	if err != nil {
+		return err
 	}
-	err = DecreaseUserQuota(relayInfo.UserId, quota)
-	return userQuota - quota, err
+	return nil
 }
 
-func PostConsumeTokenQuota(relayInfo *relaycommon.RelayInfo, userQuota int, quota int, preConsumedQuota int, sendEmail bool) (err error) {
+func PostConsumeQuota(relayInfo *relaycommon.RelayInfo, userQuota int, quota int, preConsumedQuota int, sendEmail bool) (err error) {
 
 	if quota > 0 {
 		err = DecreaseUserQuota(relayInfo.UserId, quota)
@@ -301,9 +360,9 @@ func PostConsumeTokenQuota(relayInfo *relaycommon.RelayInfo, userQuota int, quot
 
 	if !relayInfo.IsPlayground {
 		if quota > 0 {
-			err = DecreaseTokenQuota(relayInfo.TokenId, quota)
+			err = DecreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, quota)
 		} else {
-			err = IncreaseTokenQuota(relayInfo.TokenId, -quota)
+			err = IncreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, -quota)
 		}
 		if err != nil {
 			return err

model/token_cache.go

@@ -0,0 +1,64 @@
+package model
+
+import (
+	"fmt"
+	"one-api/common"
+	"one-api/constant"
+	"time"
+)
+
+func cacheSetToken(token Token) error {
+	key := common.GenerateHMAC(token.Key)
+	token.Clean()
+	err := common.RedisHSetObj(fmt.Sprintf("token:%s", key), &token, time.Duration(constant.TokenCacheSeconds)*time.Second)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+func cacheDeleteToken(key string) error {
+	key = common.GenerateHMAC(key)
+	err := common.RedisHDelObj(fmt.Sprintf("token:%s", key))
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+func cacheIncrTokenQuota(key string, increment int64) error {
+	key = common.GenerateHMAC(key)
+	err := common.RedisHIncrBy(fmt.Sprintf("token:%s", key), constant.TokenFiledRemainQuota, increment)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+func cacheDecrTokenQuota(key string, decrement int64) error {
+	return cacheIncrTokenQuota(key, -decrement)
+}
+
+func cacheSetTokenField(key string, field string, value string) error {
+	key = common.GenerateHMAC(key)
+	err := common.RedisHSetField(fmt.Sprintf("token:%s", key), field, value)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+// CacheGetTokenByKey 从缓存中获取 token，如果缓存中不存在，则从数据库中获取
+func cacheGetTokenByKey(key string) (*Token, error) {
+	hmacKey := common.GenerateHMAC(key)
+	if !common.RedisEnabled {
+		return nil, nil
+	}
+	var token Token
+	err := common.RedisHGetObj(fmt.Sprintf("token:%s", hmacKey), &token)
+	if err != nil {
+		return nil, err
+	}
+	token.Key = key
+	return &token, nil
+}

model/user.go

@@ -6,7 +6,8 @@ import (
 	"one-api/common"
 	"strconv"
 	"strings"
-	"time"
+
+	"github.com/bytedance/gopkg/util/gopool"
 
 	"gorm.io/gorm"
 )
@@ -80,46 +81,105 @@ func GetMaxUserId() int {
 	return user.Id
 }
 
-func GetAllUsers(startIdx int, num int) (users []*User, err error) {
-	err = DB.Unscoped().Order("id desc").Limit(num).Offset(startIdx).Omit("password").Find(&users).Error
-	return users, err
+func GetAllUsers(startIdx int, num int) (users []*User, total int64, err error) {
+	// Start transaction
+	tx := DB.Begin()
+	if tx.Error != nil {
+		return nil, 0, tx.Error
+	}
+	defer func() {
+		if r := recover(); r != nil {
+			tx.Rollback()
+		}
+	}()
+
+	// Get total count within transaction
+	err = tx.Unscoped().Model(&User{}).Count(&total).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// Get paginated users within same transaction
+	err = tx.Unscoped().Order("id desc").Limit(num).Offset(startIdx).Omit("password").Find(&users).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// Commit transaction
+	if err = tx.Commit().Error; err != nil {
+		return nil, 0, err
+	}
+
+	return users, total, nil
 }
 
-func SearchUsers(keyword string, group string) ([]*User, error) {
+func SearchUsers(keyword string, group string, startIdx int, num int) ([]*User, int64, error) {
 	var users []*User
+	var total int64
 	var err error
 
-	groupCol := "`group`"
-	if common.UsingPostgreSQL {
-		groupCol = `"group"`
+	// 开始事务
+	tx := DB.Begin()
+	if tx.Error != nil {
+		return nil, 0, tx.Error
 	}
+	defer func() {
+		if r := recover(); r != nil {
+			tx.Rollback()
+		}
+	}()
+
+	// 构建基础查询
+	query := tx.Unscoped().Model(&User{})
+
+	// 构建搜索条件
+	likeCondition := "username LIKE ? OR email LIKE ? OR display_name LIKE ?"
 
 	// 尝试将关键字转换为整数ID
 	keywordInt, err := strconv.Atoi(keyword)
 	if err == nil {
-		// 如果转换成功，按照ID和可选的组别搜索用户
-		query := DB.Unscoped().Omit("password").Where("id = ?", keywordInt)
+		// 如果是数字，同时搜索ID和其他字段
+		likeCondition = "id = ? OR " + likeCondition
 		if group != "" {
-			query = query.Where(groupCol+" = ?", group) // 使用反引号包围group
+			query = query.Where("("+likeCondition+") AND "+groupCol+" = ?",
+				keywordInt, "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%", group)
+		} else {
+			query = query.Where(likeCondition,
+				keywordInt, "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%")
 		}
-		err = query.Find(&users).Error
-		if err != nil || len(users) > 0 {
-			return users, err
-		}
-	}
- 
-	err = nil
-
-	query := DB.Unscoped().Omit("password")
-	likeCondition := "username LIKE ? OR email LIKE ? OR display_name LIKE ?"
-	if group != "" {
-		query = query.Where("("+likeCondition+") AND "+groupCol+" = ?", "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%", group)
 	} else {
-		query = query.Where(likeCondition, "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%")
+		// 非数字关键字，只搜索字符串字段
+		if group != "" {
+			query = query.Where("("+likeCondition+") AND "+groupCol+" = ?",
+				"%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%", group)
+		} else {
+			query = query.Where(likeCondition,
+				"%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%")
+		}
 	}
-	err = query.Find(&users).Error
 
-	return users, err
+	// 获取总数
+	err = query.Count(&total).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// 获取分页数据
+	err = query.Omit("password").Order("id desc").Limit(num).Offset(startIdx).Find(&users).Error
+	if err != nil {
+		tx.Rollback()
+		return nil, 0, err
+	}
+
+	// 提交事务
+	if err = tx.Commit().Error; err != nil {
+		return nil, 0, err
+	}
+
+	return users, total, nil
 }
 
 func GetUserById(id int, selectAll bool) (*User, error) {
@@ -251,14 +311,12 @@ func (user *User) Update(updatePassword bool) error {
 	}
 	newUser := *user
 	DB.First(&user, user.Id)
-	err = DB.Model(user).Updates(newUser).Error
-	if err == nil {
-		if common.RedisEnabled {
-			_ = common.RedisSet(fmt.Sprintf("user_group:%d", user.Id), user.Group, time.Duration(UserId2GroupCacheSeconds)*time.Second)
-			_ = common.RedisSet(fmt.Sprintf("user_quota:%d", user.Id), strconv.Itoa(user.Quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
-		}
+	if err = DB.Model(user).Updates(newUser).Error; err != nil {
+		return err
 	}
-	return err
+
+	// 更新缓存
+	return updateUserCache(user.Id, user.Username, user.Group, user.Quota, user.Status)
 }
 
 func (user *User) Edit(updatePassword bool) error {
@@ -269,6 +327,7 @@ func (user *User) Edit(updatePassword bool) error {
 			return err
 		}
 	}
+
 	newUser := *user
 	updates := map[string]interface{}{
 		"username":     newUser.Username,
@@ -279,23 +338,26 @@ func (user *User) Edit(updatePassword bool) error {
 	if updatePassword {
 		updates["password"] = newUser.Password
 	}
+
 	DB.First(&user, user.Id)
-	err = DB.Model(user).Updates(updates).Error
-	if err == nil {
-		if common.RedisEnabled {
-			_ = common.RedisSet(fmt.Sprintf("user_group:%d", user.Id), user.Group, time.Duration(UserId2GroupCacheSeconds)*time.Second)
-			_ = common.RedisSet(fmt.Sprintf("user_quota:%d", user.Id), strconv.Itoa(user.Quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
-		}
+	if err = DB.Model(user).Updates(updates).Error; err != nil {
+		return err
 	}
-	return err
+
+	// 更新缓存
+	return updateUserCache(user.Id, user.Username, user.Group, user.Quota, user.Status)
 }
 
 func (user *User) Delete() error {
 	if user.Id == 0 {
 		return errors.New("id 为空！")
 	}
-	err := DB.Delete(user).Error
-	return err
+	if err := DB.Delete(user).Error; err != nil {
+		return err
+	}
+
+	// 清除缓存
+	return invalidateUserCache(user.Id)
 }
 
 func (user *User) HardDelete() error {
@@ -409,15 +471,33 @@ func IsAdmin(userId int) bool {
 	return user.Role >= common.RoleAdminUser
 }
 
-func IsUserEnabled(userId int) (bool, error) {
-	if userId == 0 {
-		return false, errors.New("user id is empty")
+// IsUserEnabled checks user status from Redis first, falls back to DB if needed
+func IsUserEnabled(id int, fromDB bool) (status bool, err error) {
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) {
+			gopool.Go(func() {
+				if err := updateUserStatusCache(id, status); err != nil {
+					common.SysError("failed to update user status cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		// Try Redis first
+		status, err := getUserStatusCache(id)
+		if err == nil {
+			return status == common.UserStatusEnabled, nil
+		}
+		// Don't return error - fall through to DB
 	}
+	fromDB = true
 	var user User
-	err := DB.Where("id = ?", userId).Select("status").Find(&user).Error
+	err = DB.Where("id = ?", id).Select("status").Find(&user).Error
 	if err != nil {
 		return false, err
 	}
+
 	return user.Status == common.UserStatusEnabled, nil
 }
 
@@ -433,14 +513,33 @@ func ValidateAccessToken(token string) (user *User) {
 	return nil
 }
 
-func GetUserQuota(id int) (quota int, err error) {
+// GetUserQuota gets quota from Redis first, falls back to DB if needed
+func GetUserQuota(id int, fromDB bool) (quota int, err error) {
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) {
+			gopool.Go(func() {
+				if err := updateUserQuotaCache(id, quota); err != nil {
+					common.SysError("failed to update user quota cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		quota, err := getUserQuotaCache(id)
+		if err == nil {
+			return quota, nil
+		}
+		// Don't return error - fall through to DB
+		//common.SysError("failed to get user quota from cache: " + err.Error())
+	}
+	fromDB = true
 	err = DB.Model(&User{}).Where("id = ?", id).Select("quota").Find(&quota).Error
 	if err != nil {
-		if common.RedisEnabled {
-			go cacheSetUserQuota(id, quota)
-		}
+		return 0, err
 	}
-	return quota, err
+
+	return quota, nil
 }
 
 func GetUserUsedQuota(id int) (quota int, err error) {
@@ -453,20 +552,44 @@ func GetUserEmail(id int) (email string, err error) {
 	return email, err
 }
 
-func GetUserGroup(id int) (group string, err error) {
-	groupCol := "`group`"
-	if common.UsingPostgreSQL {
-		groupCol = `"group"`
+// GetUserGroup gets group from Redis first, falls back to DB if needed
+func GetUserGroup(id int, fromDB bool) (group string, err error) {
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) {
+			gopool.Go(func() {
+				if err := updateUserGroupCache(id, group); err != nil {
+					common.SysError("failed to update user group cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		group, err := getUserGroupCache(id)
+		if err == nil {
+			return group, nil
+		}
+		// Don't return error - fall through to DB
+	}
+	fromDB = true
+	err = DB.Model(&User{}).Where("id = ?", id).Select(groupCol).Find(&group).Error
+	if err != nil {
+		return "", err
 	}
 
-	err = DB.Model(&User{}).Where("id = ?", id).Select(groupCol).Find(&group).Error
-	return group, err
+	return group, nil
 }
 
 func IncreaseUserQuota(id int, quota int) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
+	gopool.Go(func() {
+		err := cacheIncrUserQuota(id, int64(quota))
+		if err != nil {
+			common.SysError("failed to increase user quota: " + err.Error())
+		}
+	})
 	if common.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeUserQuota, id, quota)
 		return nil
@@ -476,6 +599,9 @@ func IncreaseUserQuota(id int, quota int) (err error) {
 
 func increaseUserQuota(id int, quota int) (err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Update("quota", gorm.Expr("quota + ?", quota)).Error
+	if err != nil {
+		return err
+	}
 	return err
 }
 
@@ -483,6 +609,12 @@ func DecreaseUserQuota(id int, quota int) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
+	gopool.Go(func() {
+		err := cacheDecrUserQuota(id, int64(quota))
+		if err != nil {
+			common.SysError("failed to decrease user quota: " + err.Error())
+		}
+	})
 	if common.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeUserQuota, id, -quota)
 		return nil
@@ -492,9 +624,23 @@ func DecreaseUserQuota(id int, quota int) (err error) {
 
 func decreaseUserQuota(id int, quota int) (err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Update("quota", gorm.Expr("quota - ?", quota)).Error
+	if err != nil {
+		return err
+	}
 	return err
 }
 
+func DeltaUpdateUserQuota(id int, delta int) (err error) {
+	if delta == 0 {
+		return nil
+	}
+	if delta > 0 {
+		return IncreaseUserQuota(id, delta)
+	} else {
+		return DecreaseUserQuota(id, -delta)
+	}
+}
+
 func GetRootUserEmail() (email string) {
 	DB.Model(&User{}).Where("role = ?", common.RoleRootUser).Select("email").Find(&email)
 	return email
@@ -518,7 +664,13 @@ func updateUserUsedQuotaAndRequestCount(id int, quota int, count int) {
 	).Error
 	if err != nil {
 		common.SysError("failed to update user used quota and request count: " + err.Error())
+		return
 	}
+
+	//// 更新缓存
+	//if err := invalidateUserCache(id); err != nil {
+	//	common.SysError("failed to invalidate user cache: " + err.Error())
+	//}
 }
 
 func updateUserUsedQuota(id int, quota int) {
@@ -539,9 +691,32 @@ func updateUserRequestCount(id int, count int) {
 	}
 }
 
-func GetUsernameById(id int) (username string, err error) {
+// GetUsernameById gets username from Redis first, falls back to DB if needed
+func GetUsernameById(id int, fromDB bool) (username string, err error) {
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) {
+			gopool.Go(func() {
+				if err := updateUserNameCache(id, username); err != nil {
+					common.SysError("failed to update user name cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		username, err := getUserNameCache(id)
+		if err == nil {
+			return username, nil
+		}
+		// Don't return error - fall through to DB
+	}
+	fromDB = true
 	err = DB.Model(&User{}).Where("id = ?", id).Select("username").Find(&username).Error
-	return username, err
+	if err != nil {
+		return "", err
+	}
+
+	return username, nil
 }
 
 func IsLinuxDOIdAlreadyTaken(linuxDOId string) bool {

model/user_cache.go

@@ -0,0 +1,206 @@
+package model
+
+import (
+	"fmt"
+	"one-api/common"
+	"one-api/constant"
+	"strconv"
+	"time"
+)
+
+// Change UserCache struct to userCache
+type userCache struct {
+	Id       int    `json:"id"`
+	Group    string `json:"group"`
+	Quota    int    `json:"quota"`
+	Status   int    `json:"status"`
+	Role     int    `json:"role"`
+	Username string `json:"username"`
+}
+
+// Rename all exported functions to private ones
+// invalidateUserCache clears all user related cache
+func invalidateUserCache(userId int) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+
+	keys := []string{
+		fmt.Sprintf(constant.UserGroupKeyFmt, userId),
+		fmt.Sprintf(constant.UserQuotaKeyFmt, userId),
+		fmt.Sprintf(constant.UserEnabledKeyFmt, userId),
+		fmt.Sprintf(constant.UserUsernameKeyFmt, userId),
+	}
+
+	for _, key := range keys {
+		if err := common.RedisDel(key); err != nil {
+			return fmt.Errorf("failed to delete cache key %s: %w", key, err)
+		}
+	}
+	return nil
+}
+
+// updateUserGroupCache updates user group cache
+func updateUserGroupCache(userId int, group string) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	return common.RedisSet(
+		fmt.Sprintf(constant.UserGroupKeyFmt, userId),
+		group,
+		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
+	)
+}
+
+// updateUserQuotaCache updates user quota cache
+func updateUserQuotaCache(userId int, quota int) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	return common.RedisSet(
+		fmt.Sprintf(constant.UserQuotaKeyFmt, userId),
+		fmt.Sprintf("%d", quota),
+		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
+	)
+}
+
+// updateUserStatusCache updates user status cache
+func updateUserStatusCache(userId int, userEnabled bool) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	enabled := "0"
+	if userEnabled {
+		enabled = "1"
+	}
+	return common.RedisSet(
+		fmt.Sprintf(constant.UserEnabledKeyFmt, userId),
+		enabled,
+		time.Duration(constant.UserId2StatusCacheSeconds)*time.Second,
+	)
+}
+
+// updateUserNameCache updates username cache
+func updateUserNameCache(userId int, username string) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	return common.RedisSet(
+		fmt.Sprintf(constant.UserUsernameKeyFmt, userId),
+		username,
+		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
+	)
+}
+
+// updateUserCache updates all user cache fields
+func updateUserCache(userId int, username string, userGroup string, quota int, status int) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+
+	if err := updateUserGroupCache(userId, userGroup); err != nil {
+		return fmt.Errorf("update group cache: %w", err)
+	}
+
+	if err := updateUserQuotaCache(userId, quota); err != nil {
+		return fmt.Errorf("update quota cache: %w", err)
+	}
+
+	if err := updateUserStatusCache(userId, status == common.UserStatusEnabled); err != nil {
+		return fmt.Errorf("update status cache: %w", err)
+	}
+
+	if err := updateUserNameCache(userId, username); err != nil {
+		return fmt.Errorf("update username cache: %w", err)
+	}
+
+	return nil
+}
+
+// getUserGroupCache gets user group from cache
+func getUserGroupCache(userId int) (string, error) {
+	if !common.RedisEnabled {
+		return "", nil
+	}
+	return common.RedisGet(fmt.Sprintf(constant.UserGroupKeyFmt, userId))
+}
+
+// getUserQuotaCache gets user quota from cache
+func getUserQuotaCache(userId int) (int, error) {
+	if !common.RedisEnabled {
+		return 0, nil
+	}
+	quotaStr, err := common.RedisGet(fmt.Sprintf(constant.UserQuotaKeyFmt, userId))
+	if err != nil {
+		return 0, err
+	}
+	return strconv.Atoi(quotaStr)
+}
+
+// getUserStatusCache gets user status from cache
+func getUserStatusCache(userId int) (int, error) {
+	if !common.RedisEnabled {
+		return 0, nil
+	}
+	statusStr, err := common.RedisGet(fmt.Sprintf(constant.UserEnabledKeyFmt, userId))
+	if err != nil {
+		return 0, err
+	}
+	return strconv.Atoi(statusStr)
+}
+
+// getUserNameCache gets username from cache
+func getUserNameCache(userId int) (string, error) {
+	if !common.RedisEnabled {
+		return "", nil
+	}
+	return common.RedisGet(fmt.Sprintf(constant.UserUsernameKeyFmt, userId))
+}
+
+// getUserCache gets complete user cache
+func getUserCache(userId int) (*userCache, error) {
+	if !common.RedisEnabled {
+		return nil, nil
+	}
+
+	group, err := getUserGroupCache(userId)
+	if err != nil {
+		return nil, fmt.Errorf("get group cache: %w", err)
+	}
+
+	quota, err := getUserQuotaCache(userId)
+	if err != nil {
+		return nil, fmt.Errorf("get quota cache: %w", err)
+	}
+
+	status, err := getUserStatusCache(userId)
+	if err != nil {
+		return nil, fmt.Errorf("get status cache: %w", err)
+	}
+
+	username, err := getUserNameCache(userId)
+	if err != nil {
+		return nil, fmt.Errorf("get username cache: %w", err)
+	}
+
+	return &userCache{
+		Id:       userId,
+		Group:    group,
+		Quota:    quota,
+		Status:   status,
+		Username: username,
+	}, nil
+}
+
+// Add atomic quota operations
+func cacheIncrUserQuota(userId int, delta int64) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	key := fmt.Sprintf(constant.UserQuotaKeyFmt, userId)
+	return common.RedisIncr(key, delta)
+}
+
+func cacheDecrUserQuota(userId int, delta int64) error {
+	return cacheIncrUserQuota(userId, -delta)
+}

model/utils.go

@@ -88,3 +88,7 @@ func RecordExist(err error) (bool, error) {
 	}
 	return false, err
 }
+
+func shouldUpdateRedis(fromDB bool, err error) bool {
+	return common.RedisEnabled && fromDB && err == nil
+}

relay/channel/api_request.go

@@ -39,7 +39,7 @@ func DoApiRequest(a Adaptor, c *gin.Context, info *common.RelayInfo, requestBody
 	if err != nil {
 		return nil, fmt.Errorf("setup request header failed: %w", err)
 	}
-	resp, err := doRequest(c, req)
+	resp, err := doRequest(c, req, info)
 	if err != nil {
 		return nil, fmt.Errorf("do request failed: %w", err)
 	}
@@ -62,7 +62,7 @@ func DoFormRequest(a Adaptor, c *gin.Context, info *common.RelayInfo, requestBod
 	if err != nil {
 		return nil, fmt.Errorf("setup request header failed: %w", err)
 	}
-	resp, err := doRequest(c, req)
+	resp, err := doRequest(c, req, info)
 	if err != nil {
 		return nil, fmt.Errorf("do request failed: %w", err)
 	}
@@ -90,8 +90,18 @@ func DoWssRequest(a Adaptor, c *gin.Context, info *common.RelayInfo, requestBody
 	return targetConn, nil
 }
 
-func doRequest(c *gin.Context, req *http.Request) (*http.Response, error) {
-	resp, err := service.GetHttpClient().Do(req)
+func doRequest(c *gin.Context, req *http.Request, info *common.RelayInfo) (*http.Response, error) {
+	var client *http.Client
+	var err error
+	if proxyURL, ok := info.ChannelSetting["proxy"]; ok {
+		client, err = service.NewProxyHttpClient(proxyURL.(string))
+		if err != nil {
+			return nil, fmt.Errorf("new proxy http client failed: %w", err)
+		}
+	} else {
+		client = service.GetHttpClient()
+	}
+	resp, err := client.Do(req)
 	if err != nil {
 		return nil, err
 	}
@@ -120,7 +130,7 @@ func DoTaskApiRequest(a TaskAdaptor, c *gin.Context, info *common.TaskRelayInfo,
 	if err != nil {
 		return nil, fmt.Errorf("setup request header failed: %w", err)
 	}
-	resp, err := doRequest(c, req)
+	resp, err := doRequest(c, req, info.ToRelayInfo())
 	if err != nil {
 		return nil, fmt.Errorf("do request failed: %w", err)
 	}

relay/channel/aws/dto.go

@@ -10,7 +10,7 @@ type AwsClaudeRequest struct {
 	System           string                 `json:"system,omitempty"`
 	Messages         []claude.ClaudeMessage `json:"messages"`
 	MaxTokens        uint                   `json:"max_tokens,omitempty"`
-	Temperature      float64                `json:"temperature,omitempty"`
+	Temperature      *float64               `json:"temperature,omitempty"`
 	TopP             float64                `json:"top_p,omitempty"`
 	TopK             int                    `json:"top_k,omitempty"`
 	StopSequences    []string               `json:"stop_sequences,omitempty"`

relay/channel/baidu/dto.go

@@ -12,7 +12,7 @@ type BaiduMessage struct {
 
 type BaiduChatRequest struct {
 	Messages        []BaiduMessage `json:"messages"`
-	Temperature     float64        `json:"temperature,omitempty"`
+	Temperature     *float64       `json:"temperature,omitempty"`
 	TopP            float64        `json:"top_p,omitempty"`
 	PenaltyScore    float64        `json:"penalty_score,omitempty"`
 	Stream          bool           `json:"stream,omitempty"`

relay/channel/claude/dto.go

@@ -50,7 +50,7 @@ type ClaudeRequest struct {
 	MaxTokens         uint            `json:"max_tokens,omitempty"`
 	MaxTokensToSample uint            `json:"max_tokens_to_sample,omitempty"`
 	StopSequences     []string        `json:"stop_sequences,omitempty"`
-	Temperature       float64         `json:"temperature,omitempty"`
+	Temperature       *float64        `json:"temperature,omitempty"`
 	TopP              float64         `json:"top_p,omitempty"`
 	TopK              int             `json:"top_k,omitempty"`
 	//ClaudeMetadata    `json:"metadata,omitempty"`

relay/channel/claude/relay-claude.go

@@ -225,9 +225,12 @@ func RequestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*ClaudeR
 						// 判断是否是url
 						if strings.HasPrefix(imageUrl.Url, "http") {
 							// 是url，获取图片的类型和base64编码的数据
-							mimeType, data, _ := service.GetImageFromUrl(imageUrl.Url)
-							claudeMediaMessage.Source.MediaType = mimeType
-							claudeMediaMessage.Source.Data = data
+							fileData, err := service.GetFileBase64FromUrl(imageUrl.Url)
+							if err != nil {
+								return nil, fmt.Errorf("get file base64 from url failed: %s", err.Error())
+							}
+							claudeMediaMessage.Source.MediaType = fileData.MimeType
+							claudeMediaMessage.Source.Data = fileData.Base64Data
 						} else {
 							_, format, base64String, err := service.DecodeBase64ImageData(imageUrl.Url)
 							if err != nil {

relay/channel/cloudflare/dto.go

@@ -9,7 +9,7 @@ type CfRequest struct {
 	Prompt      string        `json:"prompt,omitempty"`
 	Raw         bool          `json:"raw,omitempty"`
 	Stream      bool          `json:"stream,omitempty"`
-	Temperature float64       `json:"temperature,omitempty"`
+	Temperature *float64      `json:"temperature,omitempty"`
 }
 
 type CfAudioResponse struct {

relay/channel/deepseek/constants.go

@@ -1,7 +1,7 @@
 package deepseek
 
 var ModelList = []string{
-	"deepseek-chat", "deepseek-coder",
+	"deepseek-chat", "deepseek-reasoner",
 }
 
 var ChannelName = "deepseek"

relay/channel/gemini/constant.go

@@ -3,17 +3,19 @@ package gemini
 var ModelList = []string{
 	// stable version
 	"gemini-1.5-pro", "gemini-1.5-flash", "gemini-1.5-flash-8b",
+	"gemini-2.0-flash",
 	// latest version
 	"gemini-1.5-pro-latest", "gemini-1.5-flash-latest",
-	// legacy version
-	"gemini-1.5-pro-exp-0827", "gemini-1.5-flash-exp-0827",
-	// exp
-	"gemini-exp-1114", "gemini-exp-1121", "gemini-exp-1206",
+	// preview version
+	"gemini-2.0-flash-lite-preview",
+	// gemini exp
+	"gemini-exp-1206",
 	// flash exp
 	"gemini-2.0-flash-exp",
+	// pro exp
+	"gemini-2.0-pro-exp",
 	// thinking exp
 	"gemini-2.0-flash-thinking-exp",
-	"gemini-2.0-flash-thinking-exp-1219",
 }
 
 var ChannelName = "google gemini"

relay/channel/gemini/dto.go

@@ -4,7 +4,7 @@ type GeminiChatRequest struct {
 	Contents           []GeminiChatContent        `json:"contents"`
 	SafetySettings     []GeminiChatSafetySettings `json:"safety_settings,omitempty"`
 	GenerationConfig   GeminiChatGenerationConfig `json:"generation_config,omitempty"`
-	Tools              []GeminiChatTools          `json:"tools,omitempty"`
+	Tools              []GeminiChatTool           `json:"tools,omitempty"`
 	SystemInstructions *GeminiChatContent         `json:"system_instruction,omitempty"`
 }
 
@@ -18,16 +18,39 @@ type FunctionCall struct {
 	Arguments    any    `json:"args"`
 }
 
+type GeminiFunctionResponseContent struct {
+	Name    string `json:"name"`
+	Content any    `json:"content"`
+}
+
 type FunctionResponse struct {
-	Name     string `json:"name"`
-	Response any    `json:"response"`
+	Name     string                        `json:"name"`
+	Response GeminiFunctionResponseContent `json:"response"`
+}
+
+type GeminiPartExecutableCode struct {
+	Language string `json:"language,omitempty"`
+	Code     string `json:"code,omitempty"`
+}
+
+type GeminiPartCodeExecutionResult struct {
+	Outcome string `json:"outcome,omitempty"`
+	Output  string `json:"output,omitempty"`
+}
+
+type GeminiFileData struct {
+	MimeType string `json:"mimeType,omitempty"`
+	FileUri  string `json:"fileUri,omitempty"`
 }
 
 type GeminiPart struct {
-	Text             string            `json:"text,omitempty"`
-	InlineData       *GeminiInlineData `json:"inlineData,omitempty"`
-	FunctionCall     *FunctionCall     `json:"functionCall,omitempty"`
-	FunctionResponse *FunctionResponse `json:"functionResponse,omitempty"`
+	Text                string                         `json:"text,omitempty"`
+	InlineData          *GeminiInlineData              `json:"inlineData,omitempty"`
+	FunctionCall        *FunctionCall                  `json:"functionCall,omitempty"`
+	FunctionResponse    *FunctionResponse              `json:"functionResponse,omitempty"`
+	FileData            *GeminiFileData                `json:"fileData,omitempty"`
+	ExecutableCode      *GeminiPartExecutableCode      `json:"executableCode,omitempty"`
+	CodeExecutionResult *GeminiPartCodeExecutionResult `json:"codeExecutionResult,omitempty"`
 }
 
 type GeminiChatContent struct {
@@ -40,13 +63,15 @@ type GeminiChatSafetySettings struct {
 	Threshold string `json:"threshold"`
 }
 
-type GeminiChatTools struct {
-	GoogleSearch         any `json:"googleSearch,omitempty"`
-	FunctionDeclarations any `json:"functionDeclarations,omitempty"`
+type GeminiChatTool struct {
+	GoogleSearch          any `json:"googleSearch,omitempty"`
+	GoogleSearchRetrieval any `json:"googleSearchRetrieval,omitempty"`
+	CodeExecution         any `json:"codeExecution,omitempty"`
+	FunctionDeclarations  any `json:"functionDeclarations,omitempty"`
 }
 
 type GeminiChatGenerationConfig struct {
-	Temperature      float64  `json:"temperature,omitempty"`
+	Temperature      *float64 `json:"temperature,omitempty"`
 	TopP             float64  `json:"topP,omitempty"`
 	TopK             float64  `json:"topK,omitempty"`
 	MaxOutputTokens  uint     `json:"maxOutputTokens,omitempty"`
@@ -54,11 +79,12 @@ type GeminiChatGenerationConfig struct {
 	StopSequences    []string `json:"stopSequences,omitempty"`
 	ResponseMimeType string   `json:"responseMimeType,omitempty"`
 	ResponseSchema   any      `json:"responseSchema,omitempty"`
+	Seed             int64    `json:"seed,omitempty"`
 }
 
 type GeminiChatCandidate struct {
 	Content       GeminiChatContent        `json:"content"`
-	FinishReason  string                   `json:"finishReason"`
+	FinishReason  *string                  `json:"finishReason"`
 	Index         int64                    `json:"index"`
 	SafetyRatings []GeminiChatSafetyRating `json:"safetyRatings"`
 }

relay/channel/gemini/relay-gemini.go

@@ -12,12 +12,14 @@ import (
 	relaycommon "one-api/relay/common"
 	"one-api/service"
 	"strings"
+	"unicode/utf8"
 
 	"github.com/gin-gonic/gin"
 )
 
 // Setting safety to the lowest possible values since Gemini is already powerless enough
 func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatRequest, error) {
+
 	geminiRequest := GeminiChatRequest{
 		Contents: make([]GeminiChatContent, 0, len(textRequest.Messages)),
 		SafetySettings: []GeminiChatSafetySettings{
@@ -46,16 +48,24 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 			Temperature:     textRequest.Temperature,
 			TopP:            textRequest.TopP,
 			MaxOutputTokens: textRequest.MaxTokens,
+			Seed:            int64(textRequest.Seed),
 		},
 	}
+
+	// openaiContent.FuncToToolCalls()
 	if textRequest.Tools != nil {
 		functions := make([]dto.FunctionCall, 0, len(textRequest.Tools))
 		googleSearch := false
+		codeExecution := false
 		for _, tool := range textRequest.Tools {
 			if tool.Function.Name == "googleSearch" {
 				googleSearch = true
 				continue
 			}
+			if tool.Function.Name == "codeExecution" {
+				codeExecution = true
+				continue
+			}
 			if tool.Function.Parameters != nil {
 				params, ok := tool.Function.Parameters.(map[string]interface{})
 				if ok {
@@ -68,25 +78,32 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 			}
 			functions = append(functions, tool.Function)
 		}
-		if len(functions) > 0 {
-			geminiRequest.Tools = []GeminiChatTools{
-				{
-					FunctionDeclarations: functions,
-				},
-			}
+		if codeExecution {
+			geminiRequest.Tools = append(geminiRequest.Tools, GeminiChatTool{
+				CodeExecution: make(map[string]string),
+			})
 		}
 		if googleSearch {
-			geminiRequest.Tools = append(geminiRequest.Tools, GeminiChatTools{
+			geminiRequest.Tools = append(geminiRequest.Tools, GeminiChatTool{
 				GoogleSearch: make(map[string]string),
 			})
 		}
+		if len(functions) > 0 {
+			geminiRequest.Tools = append(geminiRequest.Tools, GeminiChatTool{
+				FunctionDeclarations: functions,
+			})
+		}
+		// common.SysLog("tools: " + fmt.Sprintf("%+v", geminiRequest.Tools))
+		// json_data, _ := json.Marshal(geminiRequest.Tools)
+		// common.SysLog("tools_json: " + string(json_data))
 	} else if textRequest.Functions != nil {
-		geminiRequest.Tools = []GeminiChatTools{
+		geminiRequest.Tools = []GeminiChatTool{
 			{
 				FunctionDeclarations: textRequest.Functions,
 			},
 		}
 	}
+
 	if textRequest.ResponseFormat != nil && (textRequest.ResponseFormat.Type == "json_schema" || textRequest.ResponseFormat.Type == "json_object") {
 		geminiRequest.GenerationConfig.ResponseMimeType = "application/json"
 
@@ -96,20 +113,14 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 		}
 	}
 	tool_call_ids := make(map[string]string)
+	var system_content []string
 	//shouldAddDummyModelMessage := false
 	for _, message := range textRequest.Messages {
-
 		if message.Role == "system" {
-			geminiRequest.SystemInstructions = &GeminiChatContent{
-				Parts: []GeminiPart{
-					{
-						Text: message.StringContent(),
-					},
-				},
-			}
+			system_content = append(system_content, message.StringContent())
 			continue
-		} else if message.Role == "tool" {
-			if len(geminiRequest.Contents) == 0 || geminiRequest.Contents[len(geminiRequest.Contents)-1].Role != "user" {
+		} else if message.Role == "tool" || message.Role == "function" {
+			if len(geminiRequest.Contents) == 0 || geminiRequest.Contents[len(geminiRequest.Contents)-1].Role == "model" {
 				geminiRequest.Contents = append(geminiRequest.Contents, GeminiChatContent{
 					Role: "user",
 				})
@@ -121,9 +132,16 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 			} else if val, exists := tool_call_ids[message.ToolCallId]; exists {
 				name = val
 			}
+			content := common.StrToMap(message.StringContent())
 			functionResp := &FunctionResponse{
-				Name:     name,
-				Response: common.StrToMap(message.StringContent()),
+				Name: name,
+				Response: GeminiFunctionResponseContent{
+					Name:    name,
+					Content: content,
+				},
+			}
+			if content == nil {
+				functionResp.Response.Content = message.StringContent()
 			}
 			*parts = append(*parts, GeminiPart{
 				FunctionResponse: functionResp,
@@ -134,57 +152,68 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 		content := GeminiChatContent{
 			Role: message.Role,
 		}
-		isToolCall := false
+		// isToolCall := false
 		if message.ToolCalls != nil {
-			message.Role = "model"
-			isToolCall = true
+			// message.Role = "model"
+			// isToolCall = true
 			for _, call := range message.ParseToolCalls() {
+				args := map[string]interface{}{}
+				if call.Function.Arguments != "" {
+					if json.Unmarshal([]byte(call.Function.Arguments), &args) != nil {
+						return nil, fmt.Errorf("invalid arguments for function %s, args: %s", call.Function.Name, call.Function.Arguments)
+					}
+				}
 				toolCall := GeminiPart{
 					FunctionCall: &FunctionCall{
 						FunctionName: call.Function.Name,
-						Arguments:    call.Function.Parameters,
+						Arguments:    args,
 					},
 				}
 				parts = append(parts, toolCall)
 				tool_call_ids[call.ID] = call.Function.Name
 			}
 		}
-		if !isToolCall {
-			openaiContent := message.ParseContent()
-			imageNum := 0
-			for _, part := range openaiContent {
-				if part.Type == dto.ContentTypeText {
-					parts = append(parts, GeminiPart{
-						Text: part.Text,
-					})
-				} else if part.Type == dto.ContentTypeImageURL {
-					imageNum += 1
 
-					if constant.GeminiVisionMaxImageNum != -1 && imageNum > constant.GeminiVisionMaxImageNum {
-						return nil, fmt.Errorf("too many images in the message, max allowed is %d", constant.GeminiVisionMaxImageNum)
+		openaiContent := message.ParseContent()
+		imageNum := 0
+		for _, part := range openaiContent {
+			if part.Type == dto.ContentTypeText {
+				if part.Text == "" {
+					continue
+				}
+				parts = append(parts, GeminiPart{
+					Text: part.Text,
+				})
+			} else if part.Type == dto.ContentTypeImageURL {
+				imageNum += 1
+
+				if constant.GeminiVisionMaxImageNum != -1 && imageNum > constant.GeminiVisionMaxImageNum {
+					return nil, fmt.Errorf("too many images in the message, max allowed is %d", constant.GeminiVisionMaxImageNum)
+				}
+				// 判断是否是url
+				if strings.HasPrefix(part.ImageUrl.(dto.MessageImageUrl).Url, "http") {
+					// 是url，获取图片的类型和base64编码的数据
+					fileData, err := service.GetFileBase64FromUrl(part.ImageUrl.(dto.MessageImageUrl).Url)
+					if err != nil {
+						return nil, fmt.Errorf("get file base64 from url failed: %s", err.Error())
 					}
-					// 判断是否是url
-					if strings.HasPrefix(part.ImageUrl.(dto.MessageImageUrl).Url, "http") {
-						// 是url，获取图片的类型和base64编码的数据
-						mimeType, data, _ := service.GetImageFromUrl(part.ImageUrl.(dto.MessageImageUrl).Url)
-						parts = append(parts, GeminiPart{
-							InlineData: &GeminiInlineData{
-								MimeType: mimeType,
-								Data:     data,
-							},
-						})
-					} else {
-						_, format, base64String, err := service.DecodeBase64ImageData(part.ImageUrl.(dto.MessageImageUrl).Url)
-						if err != nil {
-							return nil, fmt.Errorf("decode base64 image data failed: %s", err.Error())
-						}
-						parts = append(parts, GeminiPart{
-							InlineData: &GeminiInlineData{
-								MimeType: "image/" + format,
-								Data:     base64String,
-							},
-						})
+					parts = append(parts, GeminiPart{
+						InlineData: &GeminiInlineData{
+							MimeType: fileData.MimeType,
+							Data:     fileData.Base64Data,
+						},
+					})
+				} else {
+					format, base64String, err := service.DecodeBase64FileData(part.ImageUrl.(dto.MessageImageUrl).Url)
+					if err != nil {
+						return nil, fmt.Errorf("decode base64 image data failed: %s", err.Error())
 					}
+					parts = append(parts, GeminiPart{
+						InlineData: &GeminiInlineData{
+							MimeType: format,
+							Data:     base64String,
+						},
+					})
 				}
 			}
 		}
@@ -197,6 +226,17 @@ func CovertGemini2OpenAI(textRequest dto.GeneralOpenAIRequest) (*GeminiChatReque
 		}
 		geminiRequest.Contents = append(geminiRequest.Contents, content)
 	}
+
+	if len(system_content) > 0 {
+		geminiRequest.SystemInstructions = &GeminiChatContent{
+			Parts: []GeminiPart{
+				{
+					Text: strings.Join(system_content, "\n"),
+				},
+			},
+		}
+	}
+
 	return &geminiRequest, nil
 }
 
@@ -209,12 +249,12 @@ func removeAdditionalPropertiesWithDepth(schema interface{}, depth int) interfac
 	if !ok || len(v) == 0 {
 		return schema
 	}
-
+	// 删除所有的title字段
+	delete(v, "title")
 	// 如果type不为object和array，则直接返回
 	if typeVal, exists := v["type"]; !exists || (typeVal != "object" && typeVal != "array") {
 		return schema
 	}
-	delete(v, "title")
 	switch v["type"] {
 	case "object":
 		delete(v, "additionalProperties")
@@ -240,20 +280,85 @@ func removeAdditionalPropertiesWithDepth(schema interface{}, depth int) interfac
 	return v
 }
 
-func (g *GeminiChatResponse) GetResponseText() string {
-	if g == nil {
-		return ""
+func unescapeString(s string) (string, error) {
+	var result []rune
+	escaped := false
+	i := 0
+
+	for i < len(s) {
+		r, size := utf8.DecodeRuneInString(s[i:]) // 正确解码UTF-8字符
+		if r == utf8.RuneError {
+			return "", fmt.Errorf("invalid UTF-8 encoding")
+		}
+
+		if escaped {
+			// 如果是转义符后的字符，检查其类型
+			switch r {
+			case '"':
+				result = append(result, '"')
+			case '\\':
+				result = append(result, '\\')
+			case '/':
+				result = append(result, '/')
+			case 'b':
+				result = append(result, '\b')
+			case 'f':
+				result = append(result, '\f')
+			case 'n':
+				result = append(result, '\n')
+			case 'r':
+				result = append(result, '\r')
+			case 't':
+				result = append(result, '\t')
+			case '\'':
+				result = append(result, '\'')
+			default:
+				// 如果遇到一个非法的转义字符，直接按原样输出
+				result = append(result, '\\', r)
+			}
+			escaped = false
+		} else {
+			if r == '\\' {
+				escaped = true // 记录反斜杠作为转义符
+			} else {
+				result = append(result, r)
+			}
+		}
+		i += size // 移动到下一个字符
 	}
-	if len(g.Candidates) > 0 && len(g.Candidates[0].Content.Parts) > 0 {
-		return g.Candidates[0].Content.Parts[0].Text
+
+	return string(result), nil
+}
+func unescapeMapOrSlice(data interface{}) interface{} {
+	switch v := data.(type) {
+	case map[string]interface{}:
+		for k, val := range v {
+			v[k] = unescapeMapOrSlice(val)
+		}
+	case []interface{}:
+		for i, val := range v {
+			v[i] = unescapeMapOrSlice(val)
+		}
+	case string:
+		if unescaped, err := unescapeString(v); err != nil {
+			return v
+		} else {
+			return unescaped
+		}
 	}
-	return ""
+	return data
 }
 
 func getToolCall(item *GeminiPart) *dto.ToolCall {
-	argsBytes, err := json.Marshal(item.FunctionCall.Arguments)
+	var argsBytes []byte
+	var err error
+	if result, ok := item.FunctionCall.Arguments.(map[string]interface{}); ok {
+		argsBytes, err = json.Marshal(unescapeMapOrSlice(result))
+	} else {
+		argsBytes, err = json.Marshal(item.FunctionCall.Arguments)
+	}
+
 	if err != nil {
-		//common.SysError("getToolCall failed: " + err.Error())
 		return nil
 	}
 	return &dto.ToolCall{
@@ -266,30 +371,6 @@ func getToolCall(item *GeminiPart) *dto.ToolCall {
 	}
 }
 
-// func getToolCalls(candidate *GeminiChatCandidate, index int) []dto.ToolCall {
-// 	var toolCalls []dto.ToolCall
-
-// 	item := candidate.Content.Parts[index]
-// 	if item.FunctionCall == nil {
-// 		return toolCalls
-// 	}
-// 	argsBytes, err := json.Marshal(item.FunctionCall.Arguments)
-// 	if err != nil {
-// 		//common.SysError("getToolCalls failed: " + err.Error())
-// 		return toolCalls
-// 	}
-// 	toolCall := dto.ToolCall{
-// 		ID:   fmt.Sprintf("call_%s", common.GetUUID()),
-// 		Type: "function",
-// 		Function: dto.FunctionCall{
-// 			Arguments: string(argsBytes),
-// 			Name:      item.FunctionCall.FunctionName,
-// 		},
-// 	}
-// 	toolCalls = append(toolCalls, toolCall)
-// 	return toolCalls
-// }
-
 func responseGeminiChat2OpenAI(response *GeminiChatResponse) *dto.OpenAITextResponse {
 	fullTextResponse := dto.OpenAITextResponse{
 		Id:      fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
@@ -298,11 +379,10 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *dto.OpenAITextResp
 		Choices: make([]dto.OpenAITextResponseChoice, 0, len(response.Candidates)),
 	}
 	content, _ := json.Marshal("")
-	for i, candidate := range response.Candidates {
-		// jsonData, _ := json.MarshalIndent(candidate, "", "  ")
-		// common.SysLog(fmt.Sprintf("candidate: %v", string(jsonData)))
+	is_tool_call := false
+	for _, candidate := range response.Candidates {
 		choice := dto.OpenAITextResponseChoice{
-			Index: i,
+			Index: int(candidate.Index),
 			Message: dto.Message{
 				Role:    "assistant",
 				Content: content,
@@ -319,48 +399,107 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *dto.OpenAITextResp
 						tool_calls = append(tool_calls, *call)
 					}
 				} else {
-					texts = append(texts, part.Text)
+					if part.ExecutableCode != nil {
+						texts = append(texts, "```"+part.ExecutableCode.Language+"\n"+part.ExecutableCode.Code+"\n```")
+					} else if part.CodeExecutionResult != nil {
+						texts = append(texts, "```output\n"+part.CodeExecutionResult.Output+"\n```")
+					} else {
+						// 过滤掉空行
+						if part.Text != "\n" {
+							texts = append(texts, part.Text)
+						}
+					}
 				}
 			}
+			if len(tool_calls) > 0 {
+				choice.Message.SetToolCalls(tool_calls)
+				is_tool_call = true
+			}
+
 			choice.Message.SetStringContent(strings.Join(texts, "\n"))
-			choice.Message.SetToolCalls(tool_calls)
+
 		}
+		if candidate.FinishReason != nil {
+			switch *candidate.FinishReason {
+			case "STOP":
+				choice.FinishReason = constant.FinishReasonStop
+			case "MAX_TOKENS":
+				choice.FinishReason = constant.FinishReasonLength
+			default:
+				choice.FinishReason = constant.FinishReasonContentFilter
+			}
+		}
+		if is_tool_call {
+			choice.FinishReason = constant.FinishReasonToolCalls
+		}
+
 		fullTextResponse.Choices = append(fullTextResponse.Choices, choice)
 	}
 	return &fullTextResponse
 }
 
-func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) *dto.ChatCompletionsStreamResponse {
-	var choice dto.ChatCompletionsStreamResponseChoice
-	//choice.Delta.SetContentString(geminiResponse.GetResponseText())
-	if len(geminiResponse.Candidates) > 0 && len(geminiResponse.Candidates[0].Content.Parts) > 0 {
+func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) (*dto.ChatCompletionsStreamResponse, bool) {
+	choices := make([]dto.ChatCompletionsStreamResponseChoice, 0, len(geminiResponse.Candidates))
+	is_stop := false
+	for _, candidate := range geminiResponse.Candidates {
+		if candidate.FinishReason != nil && *candidate.FinishReason == "STOP" {
+			is_stop = true
+			candidate.FinishReason = nil
+		}
+		choice := dto.ChatCompletionsStreamResponseChoice{
+			Index: int(candidate.Index),
+			Delta: dto.ChatCompletionsStreamResponseChoiceDelta{
+				Role: "assistant",
+			},
+		}
 		var texts []string
-		var tool_calls []dto.ToolCall
-		for _, part := range geminiResponse.Candidates[0].Content.Parts {
-			if part.FunctionCall != nil {
-				if call := getToolCall(&part); call != nil {
-					tool_calls = append(tool_calls, *call)
-				}
-			} else {
-				texts = append(texts, part.Text)
+		isTools := false
+		if candidate.FinishReason != nil {
+			// p := GeminiConvertFinishReason(*candidate.FinishReason)
+			switch *candidate.FinishReason {
+			case "STOP":
+				choice.FinishReason = &constant.FinishReasonStop
+			case "MAX_TOKENS":
+				choice.FinishReason = &constant.FinishReasonLength
+			default:
+				choice.FinishReason = &constant.FinishReasonContentFilter
 			}
 		}
-		if len(texts) > 0 {
-			choice.Delta.SetContentString(strings.Join(texts, "\n"))
+		for _, part := range candidate.Content.Parts {
+			if part.FunctionCall != nil {
+				isTools = true
+				if call := getToolCall(&part); call != nil {
+					call.SetIndex(len(choice.Delta.ToolCalls))
+					choice.Delta.ToolCalls = append(choice.Delta.ToolCalls, *call)
+				}
+			} else {
+				if part.ExecutableCode != nil {
+					texts = append(texts, "```"+part.ExecutableCode.Language+"\n"+part.ExecutableCode.Code+"\n```\n")
+				} else if part.CodeExecutionResult != nil {
+					texts = append(texts, "```output\n"+part.CodeExecutionResult.Output+"\n```\n")
+				} else {
+					if part.Text != "\n" {
+						texts = append(texts, part.Text)
+					}
+				}
+			}
 		}
-		if len(tool_calls) > 0 {
-			choice.Delta.ToolCalls = tool_calls
+		choice.Delta.SetContentString(strings.Join(texts, "\n"))
+		if isTools {
+			choice.FinishReason = &constant.FinishReasonToolCalls
 		}
+		choices = append(choices, choice)
 	}
+
 	var response dto.ChatCompletionsStreamResponse
 	response.Object = "chat.completion.chunk"
 	response.Model = "gemini"
-	response.Choices = []dto.ChatCompletionsStreamResponseChoice{choice}
-	return &response
+	response.Choices = choices
+	return &response, is_stop
 }
 
 func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
-	responseText := ""
+	// responseText := ""
 	id := fmt.Sprintf("chatcmpl-%s", common.GetUUID())
 	createAt := common.GetTimestamp()
 	var usage = &dto.Usage{}
@@ -384,14 +523,11 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 			continue
 		}
 
-		response := streamResponseGeminiChat2OpenAI(&geminiResponse)
-		if response == nil {
-			continue
-		}
+		response, is_stop := streamResponseGeminiChat2OpenAI(&geminiResponse)
 		response.Id = id
 		response.Created = createAt
 		response.Model = info.UpstreamModelName
-		responseText += response.Choices[0].Delta.GetContentString()
+		// responseText += response.Choices[0].Delta.GetContentString()
 		if geminiResponse.UsageMetadata.TotalTokenCount != 0 {
 			usage.PromptTokens = geminiResponse.UsageMetadata.PromptTokenCount
 			usage.CompletionTokens = geminiResponse.UsageMetadata.CandidatesTokenCount
@@ -400,12 +536,17 @@ func GeminiChatStreamHandler(c *gin.Context, resp *http.Response, info *relaycom
 		if err != nil {
 			common.LogError(c, err.Error())
 		}
+		if is_stop {
+			response := service.GenerateStopResponse(id, createAt, info.UpstreamModelName, constant.FinishReasonStop)
+			service.ObjectData(c, response)
+		}
 	}
 
-	response := service.GenerateStopResponse(id, createAt, info.UpstreamModelName, constant.FinishReasonStop)
-	service.ObjectData(c, response)
+	var response *dto.ChatCompletionsStreamResponse
 
 	usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
+	usage.PromptTokensDetails.TextTokens = usage.PromptTokens
+	usage.CompletionTokenDetails.TextTokens = usage.CompletionTokens
 
 	if info.ShouldIncludeUsage {
 		response = service.GenerateFinalUsageResponse(id, createAt, info.UpstreamModelName, *usage)

relay/channel/ollama/dto.go

@@ -6,7 +6,7 @@ type OllamaRequest struct {
 	Model            string         `json:"model,omitempty"`
 	Messages         []dto.Message  `json:"messages,omitempty"`
 	Stream           bool           `json:"stream,omitempty"`
-	Temperature      float64        `json:"temperature,omitempty"`
+	Temperature      *float64       `json:"temperature,omitempty"`
 	Seed             float64        `json:"seed,omitempty"`
 	Topp             float64        `json:"top_p,omitempty"`
 	TopK             int            `json:"top_k,omitempty"`
@@ -18,14 +18,14 @@ type OllamaRequest struct {
 }
 
 type Options struct {
-	Seed             int     `json:"seed,omitempty"`
-	Temperature      float64 `json:"temperature,omitempty"`
-	TopK             int     `json:"top_k,omitempty"`
-	TopP             float64 `json:"top_p,omitempty"`
-	FrequencyPenalty float64 `json:"frequency_penalty,omitempty"`
-	PresencePenalty  float64 `json:"presence_penalty,omitempty"`
-	NumPredict       int     `json:"num_predict,omitempty"`
-	NumCtx           int     `json:"num_ctx,omitempty"`
+	Seed             int      `json:"seed,omitempty"`
+	Temperature      *float64 `json:"temperature,omitempty"`
+	TopK             int      `json:"top_k,omitempty"`
+	TopP             float64  `json:"top_p,omitempty"`
+	FrequencyPenalty float64  `json:"frequency_penalty,omitempty"`
+	PresencePenalty  float64  `json:"presence_penalty,omitempty"`
+	NumPredict       int      `json:"num_predict,omitempty"`
+	NumCtx           int      `json:"num_ctx,omitempty"`
 }
 
 type OllamaEmbeddingRequest struct {

relay/channel/openai/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"mime/multipart"
 	"net/http"
 	"one-api/common"
+	constant2 "one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel"
 	"one-api/relay/channel/ai360"
@@ -44,16 +45,20 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 	}
 	switch info.ChannelType {
 	case common.ChannelTypeAzure:
+		apiVersion := info.ApiVersion
+		if apiVersion == "" {
+			apiVersion = constant2.AzureDefaultAPIVersion
+		}
 		// https://learn.microsoft.com/en-us/azure/cognitive-services/openai/chatgpt-quickstart?pivots=rest-api&tabs=command-line#rest-api
 		requestURL := strings.Split(info.RequestURLPath, "?")[0]
-		requestURL = fmt.Sprintf("%s?api-version=%s", requestURL, info.ApiVersion)
+		requestURL = fmt.Sprintf("%s?api-version=%s", requestURL, apiVersion)
 		task := strings.TrimPrefix(requestURL, "/v1/")
 		model_ := info.UpstreamModelName
 		model_ = strings.Replace(model_, ".", "", -1)
 		// https://github.com/songquanpeng/one-api/issues/67
 		requestURL = fmt.Sprintf("/openai/deployments/%s/%s", model_, task)
 		if info.RelayMode == constant.RelayModeRealtime {
-			requestURL = fmt.Sprintf("/openai/realtime?deployment=%s&api-version=%s", model_, info.ApiVersion)
+			requestURL = fmt.Sprintf("/openai/realtime?deployment=%s&api-version=%s", model_, apiVersion)
 		}
 		return relaycommon.GetFullRequestURL(info.BaseUrl, requestURL, info.ChannelType), nil
 	case common.ChannelTypeMiniMax:
@@ -106,16 +111,31 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, info *relaycommon.RelayInfo, re
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	if info.ChannelType != common.ChannelTypeOpenAI {
+	if info.ChannelType != common.ChannelTypeOpenAI && info.ChannelType != common.ChannelTypeAzure {
 		request.StreamOptions = nil
 	}
-	if strings.HasPrefix(request.Model, "o1") {
+	if strings.HasPrefix(request.Model, "o1") || strings.HasPrefix(request.Model, "o3") {
 		if request.MaxCompletionTokens == 0 && request.MaxTokens != 0 {
 			request.MaxCompletionTokens = request.MaxTokens
 			request.MaxTokens = 0
 		}
+		if strings.HasPrefix(request.Model, "o3") {
+			request.Temperature = nil
+		}
+		if strings.HasSuffix(request.Model, "-high") {
+			request.ReasoningEffort = "high"
+			request.Model = strings.TrimSuffix(request.Model, "-high")
+		} else if strings.HasSuffix(request.Model, "-low") {
+			request.ReasoningEffort = "low"
+			request.Model = strings.TrimSuffix(request.Model, "-low")
+		} else if strings.HasSuffix(request.Model, "-medium") {
+			request.ReasoningEffort = "medium"
+			request.Model = strings.TrimSuffix(request.Model, "-medium")
+		}
+		info.ReasoningEffort = request.ReasoningEffort
+		info.UpstreamModelName = request.Model
 	}
-	if request.Model == "o1" || request.Model == "o1-2024-12-17" {
+	if request.Model == "o1" || request.Model == "o1-2024-12-17" || strings.HasPrefix(request.Model, "o3") {
 		//修改第一个Message的内容，将system改为developer
 		if len(request.Messages) > 0 && request.Messages[0].Role == "system" {
 			request.Messages[0].Role = "developer"

relay/channel/openai/constant.go

@@ -13,9 +13,14 @@ var ModelList = []string{
 	"gpt-4o-mini", "gpt-4o-mini-2024-07-18",
 	"o1-preview", "o1-preview-2024-09-12",
 	"o1-mini", "o1-mini-2024-09-12",
+	"o3-mini", "o3-mini-2025-01-31",
+	"o3-mini-high", "o3-mini-2025-01-31-high",
+	"o3-mini-low", "o3-mini-2025-01-31-low",
+	"o3-mini-medium", "o3-mini-2025-01-31-medium",
 	"o1", "o1-2024-12-17",
 	"gpt-4o-audio-preview", "gpt-4o-audio-preview-2024-10-01",
-	"gpt-4o-realtime-preview", "gpt-4o-realtime-preview-2024-10-01",
+	"gpt-4o-realtime-preview", "gpt-4o-realtime-preview-2024-10-01", "gpt-4o-realtime-preview-2024-12-17",
+	"gpt-4o-mini-realtime-preview", "gpt-4o-mini-realtime-preview-2024-12-17",
 	"text-embedding-ada-002", "text-embedding-3-small", "text-embedding-3-large",
 	"text-curie-001", "text-babbage-001", "text-ada-001",
 	"text-moderation-latest", "text-moderation-stable",

relay/channel/openai/relay-openai.go

@@ -5,7 +5,10 @@ import (
 	"bytes"
 	"encoding/json"
 	"fmt"
+	"github.com/pkg/errors"
 	"io"
+	"math"
+	"mime/multipart"
 	"net/http"
 	"one-api/common"
 	"one-api/constant"
@@ -13,6 +16,7 @@ import (
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
+	"os"
 	"strings"
 	"sync"
 	"time"
@@ -65,8 +69,12 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	scanner.Split(bufio.ScanLines)
 
 	service.SetEventStreamHeaders(c)
-
-	ticker := time.NewTicker(time.Duration(constant.StreamingTimeout) * time.Second)
+	streamingTimeout := time.Duration(constant.StreamingTimeout) * time.Second
+	if strings.HasPrefix(info.UpstreamModelName, "o1") || strings.HasPrefix(info.UpstreamModelName, "o3") {
+		// twice timeout for o1 model
+		streamingTimeout *= 2
+	}
+	ticker := time.NewTicker(streamingTimeout)
 	defer ticker.Stop()
 
 	stopChan := make(chan bool)
@@ -312,6 +320,11 @@ func OpenaiTTSHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 }
 
 func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo, responseFormat string) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	// count tokens by audio file duration
+	audioTokens, err := countAudioTokens(c)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "count_audio_tokens_failed", http.StatusInternalServerError), nil
+	}
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
@@ -336,70 +349,52 @@ func OpenaiSTTHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 	}
 	resp.Body.Close()
 
-	var text string
-	switch responseFormat {
-	case "json":
-		text, err = getTextFromJSON(responseBody)
-	case "text":
-		text, err = getTextFromText(responseBody)
-	case "srt":
-		text, err = getTextFromSRT(responseBody)
-	case "verbose_json":
-		text, err = getTextFromVerboseJSON(responseBody)
-	case "vtt":
-		text, err = getTextFromVTT(responseBody)
-	}
-
 	usage := &dto.Usage{}
-	usage.PromptTokens = info.PromptTokens
-	usage.CompletionTokens, _ = service.CountTextToken(text, info.UpstreamModelName)
+	usage.PromptTokens = audioTokens
+	usage.CompletionTokens = 0
 	usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
 	return nil, usage
 }
 
-func getTextFromVTT(body []byte) (string, error) {
-	return getTextFromSRT(body)
-}
-
-func getTextFromVerboseJSON(body []byte) (string, error) {
-	var whisperResponse dto.WhisperVerboseJSONResponse
-	if err := json.Unmarshal(body, &whisperResponse); err != nil {
-		return "", fmt.Errorf("unmarshal_response_body_failed err :%w", err)
+func countAudioTokens(c *gin.Context) (int, error) {
+	body, err := common.GetRequestBody(c)
+	if err != nil {
+		return 0, errors.WithStack(err)
 	}
-	return whisperResponse.Text, nil
-}
 
-func getTextFromSRT(body []byte) (string, error) {
-	scanner := bufio.NewScanner(strings.NewReader(string(body)))
-	var builder strings.Builder
-	var textLine bool
-	for scanner.Scan() {
-		line := scanner.Text()
-		if textLine {
-			builder.WriteString(line)
-			textLine = false
-			continue
-		} else if strings.Contains(line, "-->") {
-			textLine = true
-			continue
-		}
+	var reqBody struct {
+		File *multipart.FileHeader `form:"file" binding:"required"`
 	}
-	if err := scanner.Err(); err != nil {
-		return "", err
+	c.Request.Body = io.NopCloser(bytes.NewReader(body))
+	if err = c.ShouldBind(&reqBody); err != nil {
+		return 0, errors.WithStack(err)
 	}
-	return builder.String(), nil
-}
 
-func getTextFromText(body []byte) (string, error) {
-	return strings.TrimSuffix(string(body), "\n"), nil
-}
-
-func getTextFromJSON(body []byte) (string, error) {
-	var whisperResponse dto.AudioResponse
-	if err := json.Unmarshal(body, &whisperResponse); err != nil {
-		return "", fmt.Errorf("unmarshal_response_body_failed err :%w", err)
+	reqFp, err := reqBody.File.Open()
+	if err != nil {
+		return 0, errors.WithStack(err)
 	}
-	return whisperResponse.Text, nil
+
+	tmpFp, err := os.CreateTemp("", "audio-*")
+	if err != nil {
+		return 0, errors.WithStack(err)
+	}
+	defer os.Remove(tmpFp.Name())
+
+	_, err = io.Copy(tmpFp, reqFp)
+	if err != nil {
+		return 0, errors.WithStack(err)
+	}
+	if err = tmpFp.Close(); err != nil {
+		return 0, errors.WithStack(err)
+	}
+
+	duration, err := common.GetAudioDuration(c.Request.Context(), tmpFp.Name())
+	if err != nil {
+		return 0, errors.WithStack(err)
+	}
+
+	return int(math.Round(math.Ceil(duration) / 60.0 * 1000)), nil // 1 minute 相当于 1k tokens
 }
 
 func OpenaiRealtimeHandler(c *gin.Context, info *relaycommon.RelayInfo) (*dto.OpenAIErrorWithStatusCode, *dto.RealtimeUsage) {

relay/channel/palm/dto.go

@@ -18,7 +18,7 @@ type PaLMPrompt struct {
 
 type PaLMChatRequest struct {
 	Prompt         PaLMPrompt `json:"prompt"`
-	Temperature    float64    `json:"temperature,omitempty"`
+	Temperature    *float64   `json:"temperature,omitempty"`
 	CandidateCount int        `json:"candidateCount,omitempty"`
 	TopP           float64    `json:"topP,omitempty"`
 	TopK           uint       `json:"topK,omitempty"`

relay/channel/siliconflow/constant.go

@@ -40,7 +40,7 @@ var ModelList = []string{
 	"Pro/meta-llama/Meta-Llama-3-8B-Instruct",
 	"Pro/mistralai/Mistral-7B-Instruct-v0.2",
 	"black-forest-labs/FLUX.1-schnell",
-	"iic/SenseVoiceSmall",
+	"FunAudioLLM/SenseVoiceSmall",
 	"netease-youdao/bce-embedding-base_v1",
 	"BAAI/bge-m3",
 	"internlm/internlm2_5-20b-chat",

relay/channel/tencent/relay-tencent.go

@@ -39,9 +39,7 @@ func requestOpenAI2Tencent(a *Adaptor, request dto.GeneralOpenAIRequest) *Tencen
 	if request.TopP != 0 {
 		req.TopP = &request.TopP
 	}
-	if request.Temperature != 0 {
-		req.Temperature = &request.Temperature
-	}
+	req.Temperature = request.Temperature
 	return &req
 }
 

relay/channel/vertex/dto.go

@@ -9,7 +9,7 @@ type VertexAIClaudeRequest struct {
 	MaxTokens        int                    `json:"max_tokens,omitempty"`
 	StopSequences    []string               `json:"stop_sequences,omitempty"`
 	Stream           bool                   `json:"stream,omitempty"`
-	Temperature      float64                `json:"temperature,omitempty"`
+	Temperature      *float64               `json:"temperature,omitempty"`
 	TopP             float64                `json:"top_p,omitempty"`
 	TopK             int                    `json:"top_k,omitempty"`
 	Tools            []claude.Tool          `json:"tools,omitempty"`

relay/channel/xunfei/dto.go

@@ -13,11 +13,11 @@ type XunfeiChatRequest struct {
 	} `json:"header"`
 	Parameter struct {
 		Chat struct {
-			Domain      string  `json:"domain,omitempty"`
-			Temperature float64 `json:"temperature,omitempty"`
-			TopK        int     `json:"top_k,omitempty"`
-			MaxTokens   uint    `json:"max_tokens,omitempty"`
-			Auditing    bool    `json:"auditing,omitempty"`
+			Domain      string   `json:"domain,omitempty"`
+			Temperature *float64 `json:"temperature,omitempty"`
+			TopK        int      `json:"top_k,omitempty"`
+			MaxTokens   uint     `json:"max_tokens,omitempty"`
+			Auditing    bool     `json:"auditing,omitempty"`
 		} `json:"chat"`
 	} `json:"parameter"`
 	Payload struct {

relay/channel/zhipu/dto.go

@@ -12,7 +12,7 @@ type ZhipuMessage struct {
 
 type ZhipuRequest struct {
 	Prompt      []ZhipuMessage `json:"prompt"`
-	Temperature float64        `json:"temperature,omitempty"`
+	Temperature *float64       `json:"temperature,omitempty"`
 	TopP        float64        `json:"top_p,omitempty"`
 	RequestId   string         `json:"request_id,omitempty"`
 	Incremental bool           `json:"incremental,omitempty"`

relay/common/relay_info.go

@@ -30,6 +30,7 @@ type RelayInfo struct {
 	RelayMode            int
 	UpstreamModelName    string
 	OriginModelName      string
+	RecodeModelName      string
 	RequestURLPath       string
 	ApiVersion           string
 	PromptTokens         int
@@ -45,6 +46,7 @@ type RelayInfo struct {
 	RealtimeTools        []dto.RealTimeTool
 	IsFirstRequest       bool
 	AudioUsage           bool
+	ReasoningEffort      string
 	ChannelSetting       map[string]interface{}
 }
 
@@ -87,6 +89,7 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 		FirstResponseTime: startTime.Add(-time.Second),
 		OriginModelName:   c.GetString("original_model"),
 		UpstreamModelName: c.GetString("original_model"),
+		RecodeModelName:   c.GetString("recode_model"),
 		ApiType:           apiType,
 		ApiVersion:        c.GetString("api_version"),
 		ApiKey:            strings.TrimPrefix(c.Request.Header.Get("Authorization"), "Bearer "),
@@ -109,7 +112,7 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 	}
 	if info.ChannelType == common.ChannelTypeOpenAI || info.ChannelType == common.ChannelTypeAnthropic ||
 		info.ChannelType == common.ChannelTypeAws || info.ChannelType == common.ChannelTypeGemini ||
-		info.ChannelType == common.ChannelCloudflare {
+		info.ChannelType == common.ChannelCloudflare || info.ChannelType == common.ChannelTypeAzure {
 		info.SupportStreamOptions = true
 	}
 	return info

relay/relay-audio.go

@@ -74,30 +74,16 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	}
 
 	modelRatio := common.GetModelRatio(audioRequest.Model)
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 	ratio := modelRatio * groupRatio
 	preConsumedQuota := int(float64(preConsumedTokens) * ratio)
-	userQuota, err := model.CacheGetUserQuota(relayInfo.UserId)
+	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		return service.OpenAIErrorWrapperLocal(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
-	if userQuota-preConsumedQuota < 0 {
-		return service.OpenAIErrorWrapperLocal(errors.New(fmt.Sprintf("audio pre-consumed quota failed, user quota: %d, need quota: %d", userQuota, preConsumedQuota)), "insufficient_user_quota", http.StatusBadRequest)
-	}
-	err = model.CacheDecreaseUserQuota(relayInfo.UserId, preConsumedQuota)
-	if err != nil {
-		return service.OpenAIErrorWrapperLocal(err, "decrease_user_quota_failed", http.StatusInternalServerError)
-	}
-	if userQuota > 100*preConsumedQuota {
-		// in this case, we do not pre-consume quota
-		// because the user has enough quota
-		preConsumedQuota = 0
-	}
-	if preConsumedQuota > 0 {
-		userQuota, err = model.PreConsumeTokenQuota(relayInfo, preConsumedQuota)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "pre_consume_token_quota_failed", http.StatusForbidden)
-		}
+	preConsumedQuota, userQuota, openaiErr = preConsumeQuota(c, preConsumedQuota, relayInfo)
+	if openaiErr != nil {
+		return openaiErr
 	}
 	defer func() {
 		if openaiErr != nil {

relay/relay-image.go

@@ -99,8 +99,8 @@ func ImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 		modelPrice = 0.0025 * modelRatio
 	}
 
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
-	userQuota, err := model.CacheGetUserQuota(relayInfo.UserId)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
+	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 
 	sizeRatio := 1.0
 	// Size

relay/relay-mj.go

@@ -168,9 +168,9 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 			modelPrice = defaultPrice
 		}
 	}
-	groupRatio := common.GetGroupRatio(group)
+	groupRatio := setting.GetGroupRatio(group)
 	ratio := modelPrice * groupRatio
-	userQuota, err := model.CacheGetUserQuota(userId)
+	userQuota, err := model.GetUserQuota(userId, false)
 	if err != nil {
 		return &dto.MidjourneyResponse{
 			Code:        4,
@@ -194,11 +194,11 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 	}
 	defer func(ctx context.Context) {
 		if mjResp.StatusCode == 200 && mjResp.Response.Code == 1 {
-			err := model.PostConsumeTokenQuota(relayInfo, userQuota, quota, 0, true)
+			err := model.PostConsumeQuota(relayInfo, userQuota, quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
-			err = model.CacheUpdateUserQuota(userId)
+			//err = model.CacheUpdateUserQuota(userId)
 			if err != nil {
 				common.SysError("error update user quota cache: " + err.Error())
 			}
@@ -474,9 +474,9 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 			modelPrice = defaultPrice
 		}
 	}
-	groupRatio := common.GetGroupRatio(group)
+	groupRatio := setting.GetGroupRatio(group)
 	ratio := modelPrice * groupRatio
-	userQuota, err := model.CacheGetUserQuota(userId)
+	userQuota, err := model.GetUserQuota(userId, false)
 	if err != nil {
 		return &dto.MidjourneyResponse{
 			Code:        4,
@@ -500,14 +500,10 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 
 	defer func(ctx context.Context) {
 		if consumeQuota && midjResponseWithStatus.StatusCode == 200 {
-			err := model.PostConsumeTokenQuota(relayInfo, userQuota, quota, 0, true)
+			err := model.PostConsumeQuota(relayInfo, userQuota, quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
-			err = model.CacheUpdateUserQuota(userId)
-			if err != nil {
-				common.SysError("error update user quota cache: " + err.Error())
-			}
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
 				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s，ID %s", modelPrice, groupRatio, midjRequest.Action, midjResponse.Result)

relay/relay-text.go

@@ -93,8 +93,9 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		}
 	}
 	relayInfo.UpstreamModelName = textRequest.Model
+	relayInfo.RecodeModelName = textRequest.Model
 	modelPrice, getModelPriceSuccess := common.GetModelPrice(textRequest.Model, false)
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 
 	var preConsumedQuota int
 	var ratio float64
@@ -108,10 +109,18 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		}
 	}
 
-	promptTokens, err := getPromptTokens(textRequest, relayInfo)
-	// count messages token error 计算promptTokens错误
-	if err != nil {
-		return service.OpenAIErrorWrapper(err, "count_token_messages_failed", http.StatusInternalServerError)
+	// 获取 promptTokens，如果上下文中已经存在，则直接使用
+	var promptTokens int
+	if value, exists := c.Get("prompt_tokens"); exists {
+		promptTokens = value.(int)
+		relayInfo.PromptTokens = promptTokens
+	} else {
+		promptTokens, err = getPromptTokens(textRequest, relayInfo)
+		// count messages token error 计算promptTokens错误
+		if err != nil {
+			return service.OpenAIErrorWrapper(err, "count_token_messages_failed", http.StatusInternalServerError)
+		}
+		c.Set("prompt_tokens", promptTokens)
 	}
 
 	if !getModelPriceSuccess {
@@ -210,10 +219,10 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		return openaiErr
 	}
 
-	if strings.HasPrefix(relayInfo.UpstreamModelName, "gpt-4o-audio") {
-		service.PostAudioConsumeQuota(c, relayInfo, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
+	if strings.HasPrefix(relayInfo.RecodeModelName, "gpt-4o-audio") {
+		service.PostAudioConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
 	} else {
-		postConsumeQuota(c, relayInfo, textRequest.Model, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
+		postConsumeQuota(c, relayInfo, relayInfo.RecodeModelName, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
 	}
 	return nil
 }
@@ -223,7 +232,7 @@ func getPromptTokens(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.Re
 	var err error
 	switch info.RelayMode {
 	case relayconstant.RelayModeChatCompletions:
-		promptTokens, err = service.CountTokenChatRequest(*textRequest, textRequest.Model)
+		promptTokens, err = service.CountTokenChatRequest(info, *textRequest)
 	case relayconstant.RelayModeCompletions:
 		promptTokens, err = service.CountTokenInput(textRequest.Prompt, textRequest.Model)
 	case relayconstant.RelayModeModerations:
@@ -255,7 +264,7 @@ func checkRequestSensitive(textRequest *dto.GeneralOpenAIRequest, info *relaycom
 
 // 预扣费并返回用户剩余配额
 func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommon.RelayInfo) (int, int, *dto.OpenAIErrorWithStatusCode) {
-	userQuota, err := model.CacheGetUserQuota(relayInfo.UserId)
+	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		return 0, 0, service.OpenAIErrorWrapperLocal(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
@@ -265,10 +274,6 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 	if userQuota-preConsumedQuota < 0 {
 		return 0, 0, service.OpenAIErrorWrapperLocal(fmt.Errorf("chat pre-consumed quota failed, user quota: %d, need quota: %d", userQuota, preConsumedQuota), "insufficient_user_quota", http.StatusBadRequest)
 	}
-	err = model.CacheDecreaseUserQuota(relayInfo.UserId, preConsumedQuota)
-	if err != nil {
-		return 0, 0, service.OpenAIErrorWrapperLocal(err, "decrease_user_quota_failed", http.StatusInternalServerError)
-	}
 	if userQuota > 100*preConsumedQuota {
 		// 用户额度充足，判断令牌额度是否充足
 		if !relayInfo.TokenUnlimited {
@@ -286,11 +291,16 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 			common.LogInfo(c, fmt.Sprintf("user %d with unlimited token has enough quota %d, trusted and no need to pre-consume", relayInfo.UserId, userQuota))
 		}
 	}
+
 	if preConsumedQuota > 0 {
-		userQuota, err = model.PreConsumeTokenQuota(relayInfo, preConsumedQuota)
+		err = model.PreConsumeTokenQuota(relayInfo, preConsumedQuota)
 		if err != nil {
 			return 0, 0, service.OpenAIErrorWrapperLocal(err, "pre_consume_token_quota_failed", http.StatusForbidden)
 		}
+		err = model.DecreaseUserQuota(relayInfo.UserId, preConsumedQuota)
+		if err != nil {
+			return 0, 0, service.OpenAIErrorWrapperLocal(err, "decrease_user_quota_failed", http.StatusInternalServerError)
+		}
 	}
 	return preConsumedQuota, userQuota, nil
 }
@@ -300,7 +310,7 @@ func returnPreConsumedQuota(c *gin.Context, relayInfo *relaycommon.RelayInfo, us
 		go func() {
 			relayInfoCopy := *relayInfo
 
-			err := model.PostConsumeTokenQuota(&relayInfoCopy, userQuota, -preConsumedQuota, 0, false)
+			err := model.PostConsumeQuota(&relayInfoCopy, userQuota, -preConsumedQuota, 0, false)
 			if err != nil {
 				common.SysError("error return pre-consumed quota: " + err.Error())
 			}
@@ -358,15 +368,11 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelN
 		//}
 		quotaDelta := quota - preConsumedQuota
 		if quotaDelta != 0 {
-			err := model.PostConsumeTokenQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
+			err := model.PostConsumeQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
 			if err != nil {
 				common.LogError(ctx, "error consuming token remain quota: "+err.Error())
 			}
 		}
-		err := model.CacheUpdateUserQuota(relayInfo.UserId)
-		if err != nil {
-			common.LogError(ctx, "error update user quota cache: "+err.Error())
-		}
 		model.UpdateUserUsedQuotaAndRequestCount(relayInfo.UserId, quota)
 		model.UpdateChannelUsedQuota(relayInfo.ChannelId, quota)
 	}

relay/relay_rerank.go

@@ -10,6 +10,7 @@ import (
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
+	"one-api/setting"
 )
 
 func getRerankPromptToken(rerankRequest dto.RerankRequest) int {
@@ -57,7 +58,7 @@ func RerankHelper(c *gin.Context, relayMode int) (openaiErr *dto.OpenAIErrorWith
 
 	relayInfo.UpstreamModelName = rerankRequest.Model
 	modelPrice, success := common.GetModelPrice(rerankRequest.Model, false)
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 
 	var preConsumedQuota int
 	var ratio float64

relay/relay_task.go

@@ -16,6 +16,7 @@ import (
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
+	"one-api/setting"
 )
 
 /*
@@ -48,9 +49,9 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 	}
 
 	// 预扣
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 	ratio := modelPrice * groupRatio
-	userQuota, err := model.CacheGetUserQuota(relayInfo.UserId)
+	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		taskErr = service.TaskErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
 		return
@@ -112,14 +113,10 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 		// release quota
 		if relayInfo.ConsumeQuota && taskErr == nil {
 
-			err := model.PostConsumeTokenQuota(relayInfo.ToRelayInfo(), userQuota, quota, 0, true)
+			err := model.PostConsumeQuota(relayInfo.ToRelayInfo(), userQuota, quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
-			err = model.CacheUpdateUserQuota(relayInfo.UserId)
-			if err != nil {
-				common.SysError("error update user quota cache: " + err.Error())
-			}
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
 				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, groupRatio, relayInfo.Action)

relay/websocket.go

@@ -10,26 +10,9 @@ import (
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
+	"one-api/setting"
 )
 
-//func getAndValidateWssRequest(c *gin.Context, ws *websocket.Conn) (*dto.RealtimeEvent, error) {
-//	_, p, err := ws.ReadMessage()
-//	if err != nil {
-//		return nil, err
-//	}
-//	realtimeEvent := &dto.RealtimeEvent{}
-//	err = json.Unmarshal(p, realtimeEvent)
-//	if err != nil {
-//		return nil, err
-//	}
-//	// save the original request
-//	if realtimeEvent.Session == nil {
-//		return nil, errors.New("session object is nil")
-//	}
-//	c.Set("first_wss_request", p)
-//	return realtimeEvent, nil
-//}
-
 func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 	relayInfo := relaycommon.GenRelayInfoWs(c, ws)
 
@@ -57,7 +40,7 @@ func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWi
 	}
 	//relayInfo.UpstreamModelName = textRequest.Model
 	modelPrice, getModelPriceSuccess := common.GetModelPrice(relayInfo.UpstreamModelName, false)
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 
 	var preConsumedQuota int
 	var ratio float64
@@ -128,32 +111,7 @@ func WssHelper(c *gin.Context, ws *websocket.Conn) (openaiErr *dto.OpenAIErrorWi
 		service.ResetStatusCode(openaiErr, statusCodeMappingStr)
 		return openaiErr
 	}
-	service.PostWssConsumeQuota(c, relayInfo, relayInfo.UpstreamModelName, usage.(*dto.RealtimeUsage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
+	service.PostWssConsumeQuota(c, relayInfo, relayInfo.UpstreamModelName, usage.(*dto.RealtimeUsage), preConsumedQuota,
+		userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
 	return nil
 }
-
-//func getWssPromptTokens(textRequest *dto.RealtimeEvent, info *relaycommon.RelayInfo) (int, error) {
-//	var promptTokens int
-//	var err error
-//	switch info.RelayMode {
-//	default:
-//		promptTokens, err = service.CountTokenRealtime(*textRequest, info.UpstreamModelName)
-//	}
-//	info.PromptTokens = promptTokens
-//	return promptTokens, err
-//}
-
-//func checkWssRequestSensitive(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.RelayInfo) error {
-//	var err error
-//	switch info.RelayMode {
-//	case relayconstant.RelayModeChatCompletions:
-//		err = service.CheckSensitiveMessages(textRequest.Messages)
-//	case relayconstant.RelayModeCompletions:
-//		err = service.CheckSensitiveInput(textRequest.Prompt)
-//	case relayconstant.RelayModeModerations:
-//		err = service.CheckSensitiveInput(textRequest.Input)
-//	case relayconstant.RelayModeEmbeddings:
-//		err = service.CheckSensitiveInput(textRequest.Input)
-//	}
-//	return err
-//}

router/api-router.go

@@ -28,10 +28,10 @@ func SetApiRouter(router *gin.Engine) {
 		apiRouter.GET("/oauth/linuxdo", middleware.CriticalRateLimit(), controller.LinuxdoOAuth)
 		apiRouter.GET("/oauth/state", middleware.CriticalRateLimit(), controller.GenerateOAuthCode)
 		apiRouter.GET("/oauth/wechat", middleware.CriticalRateLimit(), controller.WeChatAuth)
-		apiRouter.GET("/oauth/wechat/bind", middleware.CriticalRateLimit(), middleware.UserAuth(), controller.WeChatBind)
-		apiRouter.GET("/oauth/email/bind", middleware.CriticalRateLimit(), middleware.UserAuth(), controller.EmailBind)
+		apiRouter.GET("/oauth/wechat/bind", middleware.CriticalRateLimit(), controller.WeChatBind)
+		apiRouter.GET("/oauth/email/bind", middleware.CriticalRateLimit(), controller.EmailBind)
 		apiRouter.GET("/oauth/telegram/login", middleware.CriticalRateLimit(), controller.TelegramLogin)
-		apiRouter.GET("/oauth/telegram/bind", middleware.CriticalRateLimit(), middleware.UserAuth(), controller.TelegramBind)
+		apiRouter.GET("/oauth/telegram/bind", middleware.CriticalRateLimit(), controller.TelegramBind)
 
 		userRoute := apiRouter.Group("/user")
 		{
@@ -99,7 +99,7 @@ func SetApiRouter(router *gin.Engine) {
 			channelRoute.POST("/fix", controller.FixChannelsAbilities)
 			channelRoute.GET("/fetch_models/:id", controller.FetchUpstreamModels)
 			channelRoute.POST("/fetch_models", controller.FetchModels)
-
+			channelRoute.POST("/batch/tag", controller.BatchSetChannelTag)
 		}
 		tokenRoute := apiRouter.Group("/token")
 		tokenRoute.Use(middleware.UserAuth())

service/cf_worker.go

@@ -9,9 +9,12 @@ import (
 	"strings"
 )
 
-func DoImageRequest(originUrl string) (resp *http.Response, err error) {
+func DoDownloadRequest(originUrl string) (resp *http.Response, err error) {
 	if setting.EnableWorker() {
-		common.SysLog(fmt.Sprintf("downloading image from worker: %s", originUrl))
+		common.SysLog(fmt.Sprintf("downloading file from worker: %s", originUrl))
+		if !strings.HasPrefix(originUrl, "https") {
+			return nil, fmt.Errorf("only support https url")
+		}
 		workerUrl := setting.WorkerUrl
 		if !strings.HasSuffix(workerUrl, "/") {
 			workerUrl += "/"
@@ -20,7 +23,7 @@ func DoImageRequest(originUrl string) (resp *http.Response, err error) {
 		data := []byte(`{"url":"` + originUrl + `","key":"` + setting.WorkerValidKey + `"}`)
 		return http.Post(setting.WorkerUrl, "application/json", bytes.NewBuffer(data))
 	} else {
-		common.SysLog(fmt.Sprintf("downloading image from origin: %s", originUrl))
+		common.SysLog(fmt.Sprintf("downloading from origin: %s", originUrl))
 		return http.Get(originUrl)
 	}
 }

service/file_decoder.go

@@ -0,0 +1,39 @@
+package service
+
+import (
+	"encoding/base64"
+	"fmt"
+	"io"
+	"one-api/constant"
+	"one-api/dto"
+)
+
+var maxFileSize = constant.MaxFileDownloadMB * 1024 * 1024
+
+func GetFileBase64FromUrl(url string) (*dto.LocalFileData, error) {
+	resp, err := DoDownloadRequest(url)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+
+	// Always use LimitReader to prevent oversized downloads
+	fileBytes, err := io.ReadAll(io.LimitReader(resp.Body, int64(maxFileSize+1)))
+	if err != nil {
+		return nil, err
+	}
+
+	// Check actual size after reading
+	if len(fileBytes) > maxFileSize {
+		return nil, fmt.Errorf("file size exceeds maximum allowed size: %dMB", constant.MaxFileDownloadMB)
+	}
+
+	// Convert to base64
+	base64Data := base64.StdEncoding.EncodeToString(fileBytes)
+
+	return &dto.LocalFileData{
+		Base64Data: base64Data,
+		MimeType:   resp.Header.Get("Content-Type"),
+		Size:       int64(len(fileBytes)),
+	}, nil
+}

service/http_client.go

@@ -1,7 +1,12 @@
 package service
 
 import (
+	"context"
+	"fmt"
+	"golang.org/x/net/proxy"
+	"net"
 	"net/http"
+	"net/url"
 	"one-api/common"
 	"time"
 )
@@ -30,3 +35,54 @@ func GetHttpClient() *http.Client {
 func GetImpatientHttpClient() *http.Client {
 	return impatientHTTPClient
 }
+
+// NewProxyHttpClient 创建支持代理的 HTTP 客户端
+func NewProxyHttpClient(proxyURL string) (*http.Client, error) {
+	if proxyURL == "" {
+		return http.DefaultClient, nil
+	}
+
+	parsedURL, err := url.Parse(proxyURL)
+	if err != nil {
+		return nil, err
+	}
+
+	switch parsedURL.Scheme {
+	case "http", "https":
+		return &http.Client{
+			Transport: &http.Transport{
+				Proxy: http.ProxyURL(parsedURL),
+			},
+		}, nil
+
+	case "socks5":
+		// 获取认证信息
+		var auth *proxy.Auth
+		if parsedURL.User != nil {
+			auth = &proxy.Auth{
+				User:     parsedURL.User.Username(),
+				Password: "",
+			}
+			if password, ok := parsedURL.User.Password(); ok {
+				auth.Password = password
+			}
+		}
+
+		// 创建 SOCKS5 代理拨号器
+		dialer, err := proxy.SOCKS5("tcp", parsedURL.Host, auth, proxy.Direct)
+		if err != nil {
+			return nil, err
+		}
+
+		return &http.Client{
+			Transport: &http.Transport{
+				DialContext: func(ctx context.Context, network, addr string) (net.Conn, error) {
+					return dialer.Dial(network, addr)
+				},
+			},
+		}, nil
+
+	default:
+		return nil, fmt.Errorf("unsupported proxy scheme: %s", parsedURL.Scheme)
+	}
+}

service/image.go

@@ -5,11 +5,12 @@ import (
 	"encoding/base64"
 	"errors"
 	"fmt"
-	"golang.org/x/image/webp"
 	"image"
 	"io"
 	"one-api/common"
 	"strings"
+
+	"golang.org/x/image/webp"
 )
 
 func DecodeBase64ImageData(base64String string) (image.Config, string, string, error) {
@@ -31,14 +32,39 @@ func DecodeBase64ImageData(base64String string) (image.Config, string, string, e
 	return config, format, base64String, err
 }
 
+func DecodeBase64FileData(base64String string) (string, string, error) {
+	var mimeType string
+	var idx int
+	idx = strings.Index(base64String, ",")
+	if idx == -1 {
+		_, file_type, base64, err := DecodeBase64ImageData(base64String)
+		return "image/" + file_type, base64, err
+	}
+	mimeType = base64String[:idx]
+	base64String = base64String[idx+1:]
+	idx = strings.Index(mimeType, ";")
+	if idx == -1 {
+		_, file_type, base64, err := DecodeBase64ImageData(base64String)
+		return "image/" + file_type, base64, err
+	}
+	mimeType = mimeType[:idx]
+	idx = strings.Index(mimeType, ":")
+	if idx == -1 {
+		_, file_type, base64, err := DecodeBase64ImageData(base64String)
+		return "image/" + file_type, base64, err
+	}
+	mimeType = mimeType[idx+1:]
+	return mimeType, base64String, nil
+}
+
 // GetImageFromUrl 获取图片的类型和base64编码的数据
 func GetImageFromUrl(url string) (mimeType string, data string, err error) {
-	resp, err := DoImageRequest(url)
+	resp, err := DoDownloadRequest(url)
 	if err != nil {
-		return
+		return "", "", err
 	}
 	if !strings.HasPrefix(resp.Header.Get("Content-Type"), "image/") {
-		return
+		return "", "", fmt.Errorf("invalid content type: %s, required image/*", resp.Header.Get("Content-Type"))
 	}
 	defer resp.Body.Close()
 	buffer := bytes.NewBuffer(nil)
@@ -52,7 +78,7 @@ func GetImageFromUrl(url string) (mimeType string, data string, err error) {
 }
 
 func DecodeUrlImageData(imageUrl string) (image.Config, string, error) {
-	response, err := DoImageRequest(imageUrl)
+	response, err := DoDownloadRequest(imageUrl)
 	if err != nil {
 		common.SysLog(fmt.Sprintf("fail to get image from url: %s", err.Error()))
 		return image.Config{}, "", err
@@ -64,6 +90,12 @@ func DecodeUrlImageData(imageUrl string) (image.Config, string, error) {
 		return image.Config{}, "", err
 	}
 
+	mimeType := response.Header.Get("Content-Type")
+
+	if !strings.HasPrefix(mimeType, "image/") {
+		return image.Config{}, "", fmt.Errorf("invalid content type: %s, required image/*", mimeType)
+	}
+
 	var readData []byte
 	for _, limit := range []int64{1024 * 8, 1024 * 24, 1024 * 64} {
 		common.SysLog(fmt.Sprintf("try to decode image config with limit: %d", limit))

service/log_info_generate.go

@@ -13,6 +13,9 @@ func GenerateTextOtherInfo(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, m
 	other["completion_ratio"] = completionRatio
 	other["model_price"] = modelPrice
 	other["frt"] = float64(relayInfo.FirstResponseTime.UnixMilli() - relayInfo.StartTime.UnixMilli())
+	if relayInfo.ReasoningEffort != "" {
+		other["reasoning_effort"] = relayInfo.ReasoningEffort
+	}
 	adminInfo := make(map[string]interface{})
 	adminInfo["use_channel"] = ctx.GetStringSlice("use_channel")
 	other["admin_info"] = adminInfo

service/quota.go

@@ -3,26 +3,65 @@ package service
 import (
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"math"
 	"one-api/common"
 	"one-api/dto"
 	"one-api/model"
 	relaycommon "one-api/relay/common"
+	"one-api/setting"
 	"strings"
 	"time"
+
+	"github.com/gin-gonic/gin"
 )
 
+type TokenDetails struct {
+	TextTokens  int
+	AudioTokens int
+}
+
+type QuotaInfo struct {
+	InputDetails  TokenDetails
+	OutputDetails TokenDetails
+	ModelName     string
+	UsePrice      bool
+	ModelPrice    float64
+	ModelRatio    float64
+	GroupRatio    float64
+}
+
+func calculateAudioQuota(info QuotaInfo) int {
+	if info.UsePrice {
+		return int(info.ModelPrice * common.QuotaPerUnit * info.GroupRatio)
+	}
+
+	completionRatio := common.GetCompletionRatio(info.ModelName)
+	audioRatio := common.GetAudioRatio(info.ModelName)
+	audioCompletionRatio := common.GetAudioCompletionRatio(info.ModelName)
+	ratio := info.GroupRatio * info.ModelRatio
+
+	quota := info.InputDetails.TextTokens + int(math.Round(float64(info.OutputDetails.TextTokens)*completionRatio))
+	quota += int(math.Round(float64(info.InputDetails.AudioTokens)*audioRatio)) +
+		int(math.Round(float64(info.OutputDetails.AudioTokens)*audioRatio*audioCompletionRatio))
+
+	quota = int(math.Round(float64(quota) * ratio))
+	if ratio != 0 && quota <= 0 {
+		quota = 1
+	}
+
+	return quota
+}
+
 func PreWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usage *dto.RealtimeUsage) error {
 	if relayInfo.UsePrice {
 		return nil
 	}
-	userQuota, err := model.GetUserQuota(relayInfo.UserId)
+	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		return err
 	}
 
-	token, err := model.CacheGetTokenByKey(strings.TrimLeft(relayInfo.TokenKey, "sk-"))
+	token, err := model.GetTokenByKey(strings.TrimLeft(relayInfo.TokenKey, "sk-"), false)
 	if err != nil {
 		return err
 	}
@@ -32,23 +71,26 @@ func PreWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usag
 	textOutTokens := usage.OutputTokenDetails.TextTokens
 	audioInputTokens := usage.InputTokenDetails.AudioTokens
 	audioOutTokens := usage.OutputTokenDetails.AudioTokens
-
-	completionRatio := common.GetCompletionRatio(modelName)
-	audioRatio := common.GetAudioRatio(relayInfo.UpstreamModelName)
-	audioCompletionRatio := common.GetAudioCompletionRatio(modelName)
-	groupRatio := common.GetGroupRatio(relayInfo.Group)
+	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 	modelRatio := common.GetModelRatio(modelName)
 
-	ratio := groupRatio * modelRatio
-
-	quota := textInputTokens + int(math.Round(float64(textOutTokens)*completionRatio))
-	quota += int(math.Round(float64(audioInputTokens)*audioRatio)) + int(math.Round(float64(audioOutTokens)*audioRatio*audioCompletionRatio))
-
-	quota = int(math.Round(float64(quota) * ratio))
-	if ratio != 0 && quota <= 0 {
-		quota = 1
+	quotaInfo := QuotaInfo{
+		InputDetails: TokenDetails{
+			TextTokens:  textInputTokens,
+			AudioTokens: audioInputTokens,
+		},
+		OutputDetails: TokenDetails{
+			TextTokens:  textOutTokens,
+			AudioTokens: audioOutTokens,
+		},
+		ModelName:  modelName,
+		UsePrice:   relayInfo.UsePrice,
+		ModelRatio: modelRatio,
+		GroupRatio: groupRatio,
 	}
 
+	quota := calculateAudioQuota(quotaInfo)
+
 	if userQuota < quota {
 		return errors.New(fmt.Sprintf("用户额度不足，剩余额度为 %d", userQuota))
 	}
@@ -57,21 +99,16 @@ func PreWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usag
 		return errors.New(fmt.Sprintf("令牌额度不足，剩余额度为 %d", token.RemainQuota))
 	}
 
-	err = model.PostConsumeTokenQuota(relayInfo, 0, quota, 0, false)
+	err = model.PostConsumeQuota(relayInfo, 0, quota, 0, false)
 	if err != nil {
 		return err
 	}
 	common.LogInfo(ctx, "realtime streaming consume quota success, quota: "+fmt.Sprintf("%d", quota))
-	err = model.CacheUpdateUserQuota(relayInfo.UserId)
-	if err != nil {
-		return err
-	}
 	return nil
 }
 
 func PostWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelName string,
-	usage *dto.RealtimeUsage, ratio float64, preConsumedQuota int, userQuota int, modelRatio float64,
-	groupRatio float64,
+	usage *dto.RealtimeUsage, preConsumedQuota int, userQuota int, modelRatio float64, groupRatio float64,
 	modelPrice float64, usePrice bool, extraContent string) {
 
 	useTimeSeconds := time.Now().Unix() - relayInfo.StartTime.Unix()
@@ -86,17 +123,23 @@ func PostWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, mod
 	audioRatio := common.GetAudioRatio(relayInfo.UpstreamModelName)
 	audioCompletionRatio := common.GetAudioCompletionRatio(modelName)
 
-	quota := 0
-	if !usePrice {
-		quota = int(math.Round(float64(textInputTokens) + float64(textOutTokens)*completionRatio))
-		quota += int(math.Round(float64(audioInputTokens)*audioRatio + float64(audioOutTokens)*audioRatio*audioCompletionRatio))
-		quota = int(math.Round(float64(quota) * ratio))
-		if ratio != 0 && quota <= 0 {
-			quota = 1
-		}
-	} else {
-		quota = int(modelPrice * common.QuotaPerUnit * groupRatio)
+	quotaInfo := QuotaInfo{
+		InputDetails: TokenDetails{
+			TextTokens:  textInputTokens,
+			AudioTokens: audioInputTokens,
+		},
+		OutputDetails: TokenDetails{
+			TextTokens:  textOutTokens,
+			AudioTokens: audioOutTokens,
+		},
+		ModelName:  modelName,
+		UsePrice:   usePrice,
+		ModelRatio: modelRatio,
+		GroupRatio: groupRatio,
 	}
+
+	quota := calculateAudioQuota(quotaInfo)
+
 	totalTokens := usage.TotalTokens
 	var logContent string
 	if !usePrice {
@@ -114,21 +157,6 @@ func PostWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, mod
 		common.LogError(ctx, fmt.Sprintf("total tokens is 0, cannot consume quota, userId %d, channelId %d, "+
 			"tokenId %d, model %s， pre-consumed quota %d", relayInfo.UserId, relayInfo.ChannelId, relayInfo.TokenId, modelName, preConsumedQuota))
 	} else {
-		//if sensitiveResp != nil {
-		//	logContent += fmt.Sprintf("，敏感词：%s", strings.Join(sensitiveResp.SensitiveWords, ", "))
-		//}
-		//quotaDelta := quota - preConsumedQuota
-		//if quotaDelta != 0 {
-		//	err := model.PostConsumeTokenQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
-		//	if err != nil {
-		//		common.LogError(ctx, "error consuming token remain quota: "+err.Error())
-		//	}
-		//}
-
-		//err := model.CacheUpdateUserQuota(relayInfo.UserId)
-		//if err != nil {
-		//	common.LogError(ctx, "error update user quota cache: "+err.Error())
-		//}
 		model.UpdateUserUsedQuotaAndRequestCount(relayInfo.UserId, quota)
 		model.UpdateChannelUsedQuota(relayInfo.ChannelId, quota)
 	}
@@ -143,8 +171,7 @@ func PostWssConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, mod
 }
 
 func PostAudioConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
-	usage *dto.Usage, ratio float64, preConsumedQuota int, userQuota int, modelRatio float64,
-	groupRatio float64,
+	usage *dto.Usage, preConsumedQuota int, userQuota int, modelRatio float64, groupRatio float64,
 	modelPrice float64, usePrice bool, extraContent string) {
 
 	useTimeSeconds := time.Now().Unix() - relayInfo.StartTime.Unix()
@@ -155,21 +182,27 @@ func PostAudioConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 	audioOutTokens := usage.CompletionTokenDetails.AudioTokens
 
 	tokenName := ctx.GetString("token_name")
-	completionRatio := common.GetCompletionRatio(relayInfo.UpstreamModelName)
-	audioRatio := common.GetAudioRatio(relayInfo.UpstreamModelName)
-	audioCompletionRatio := common.GetAudioCompletionRatio(relayInfo.UpstreamModelName)
+	completionRatio := common.GetCompletionRatio(relayInfo.RecodeModelName)
+	audioRatio := common.GetAudioRatio(relayInfo.RecodeModelName)
+	audioCompletionRatio := common.GetAudioCompletionRatio(relayInfo.RecodeModelName)
 
-	quota := 0
-	if !usePrice {
-		quota = int(math.Round(float64(textInputTokens) + float64(textOutTokens)*completionRatio))
-		quota += int(math.Round(float64(audioInputTokens)*audioRatio + float64(audioOutTokens)*audioRatio*audioCompletionRatio))
-		quota = int(math.Round(float64(quota) * ratio))
-		if ratio != 0 && quota <= 0 {
-			quota = 1
-		}
-	} else {
-		quota = int(modelPrice * common.QuotaPerUnit * groupRatio)
+	quotaInfo := QuotaInfo{
+		InputDetails: TokenDetails{
+			TextTokens:  textInputTokens,
+			AudioTokens: audioInputTokens,
+		},
+		OutputDetails: TokenDetails{
+			TextTokens:  textOutTokens,
+			AudioTokens: audioOutTokens,
+		},
+		ModelName:  relayInfo.RecodeModelName,
+		UsePrice:   usePrice,
+		ModelRatio: modelRatio,
+		GroupRatio: groupRatio,
 	}
+
+	quota := calculateAudioQuota(quotaInfo)
+
 	totalTokens := usage.TotalTokens
 	var logContent string
 	if !usePrice {
@@ -185,24 +218,20 @@ func PostAudioConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
 		quota = 0
 		logContent += fmt.Sprintf("（可能是上游超时）")
 		common.LogError(ctx, fmt.Sprintf("total tokens is 0, cannot consume quota, userId %d, channelId %d, "+
-			"tokenId %d, model %s， pre-consumed quota %d", relayInfo.UserId, relayInfo.ChannelId, relayInfo.TokenId, relayInfo.UpstreamModelName, preConsumedQuota))
+			"tokenId %d, model %s， pre-consumed quota %d", relayInfo.UserId, relayInfo.ChannelId, relayInfo.TokenId, relayInfo.RecodeModelName, preConsumedQuota))
 	} else {
 		quotaDelta := quota - preConsumedQuota
 		if quotaDelta != 0 {
-			err := model.PostConsumeTokenQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
+			err := model.PostConsumeQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
 			if err != nil {
 				common.LogError(ctx, "error consuming token remain quota: "+err.Error())
 			}
 		}
-		err := model.CacheUpdateUserQuota(relayInfo.UserId)
-		if err != nil {
-			common.LogError(ctx, "error update user quota cache: "+err.Error())
-		}
 		model.UpdateUserUsedQuotaAndRequestCount(relayInfo.UserId, quota)
 		model.UpdateChannelUsedQuota(relayInfo.ChannelId, quota)
 	}
 
-	logModel := relayInfo.UpstreamModelName
+	logModel := relayInfo.RecodeModelName
 	if extraContent != "" {
 		logContent += ", " + extraContent
 	}

service/token_counter.go

@@ -19,42 +19,40 @@ import (
 // tokenEncoderMap won't grow after initialization
 var tokenEncoderMap = map[string]*tiktoken.Tiktoken{}
 var defaultTokenEncoder *tiktoken.Tiktoken
-var cl200kTokenEncoder *tiktoken.Tiktoken
+var o200kTokenEncoder *tiktoken.Tiktoken
 
 func InitTokenEncoders() {
 	common.SysLog("initializing token encoders")
-	gpt35TokenEncoder, err := tiktoken.EncodingForModel("gpt-3.5-turbo")
+	cl100TokenEncoder, err := tiktoken.GetEncoding(tiktoken.MODEL_CL100K_BASE)
 	if err != nil {
 		common.FatalLog(fmt.Sprintf("failed to get gpt-3.5-turbo token encoder: %s", err.Error()))
 	}
-	defaultTokenEncoder = gpt35TokenEncoder
-	gpt4TokenEncoder, err := tiktoken.EncodingForModel("gpt-4")
-	if err != nil {
-		common.FatalLog(fmt.Sprintf("failed to get gpt-4 token encoder: %s", err.Error()))
-	}
-	cl200kTokenEncoder, err = tiktoken.EncodingForModel("gpt-4o")
+	defaultTokenEncoder = cl100TokenEncoder
+	o200kTokenEncoder, err = tiktoken.GetEncoding(tiktoken.MODEL_O200K_BASE)
 	if err != nil {
 		common.FatalLog(fmt.Sprintf("failed to get gpt-4o token encoder: %s", err.Error()))
 	}
 	for model, _ := range common.GetDefaultModelRatioMap() {
 		if strings.HasPrefix(model, "gpt-3.5") {
-			tokenEncoderMap[model] = gpt35TokenEncoder
+			tokenEncoderMap[model] = cl100TokenEncoder
 		} else if strings.HasPrefix(model, "gpt-4") {
 			if strings.HasPrefix(model, "gpt-4o") {
-				tokenEncoderMap[model] = cl200kTokenEncoder
+				tokenEncoderMap[model] = o200kTokenEncoder
 			} else {
-				tokenEncoderMap[model] = gpt4TokenEncoder
+				tokenEncoderMap[model] = defaultTokenEncoder
 			}
+		} else if strings.HasPrefix(model, "o1") {
+			tokenEncoderMap[model] = o200kTokenEncoder
 		} else {
-			tokenEncoderMap[model] = nil
+			tokenEncoderMap[model] = defaultTokenEncoder
 		}
 	}
 	common.SysLog("token encoders initialized")
 }
 
 func getModelDefaultTokenEncoder(model string) *tiktoken.Tiktoken {
-	if strings.HasPrefix(model, "gpt-4o") || strings.HasPrefix(model, "chatgpt-4o") {
-		return cl200kTokenEncoder
+	if strings.HasPrefix(model, "gpt-4o") || strings.HasPrefix(model, "chatgpt-4o") || strings.HasPrefix(model, "o1") {
+		return o200kTokenEncoder
 	}
 	return defaultTokenEncoder
 }
@@ -82,7 +80,7 @@ func getTokenNum(tokenEncoder *tiktoken.Tiktoken, text string) int {
 	return len(tokenEncoder.Encode(text, nil, nil))
 }
 
-func getImageToken(imageUrl *dto.MessageImageUrl, model string, stream bool) (int, error) {
+func getImageToken(info *relaycommon.RelayInfo, imageUrl *dto.MessageImageUrl, model string, stream bool) (int, error) {
 	baseTokens := 85
 	if model == "glm-4v" {
 		return 1047, nil
@@ -92,11 +90,7 @@ func getImageToken(imageUrl *dto.MessageImageUrl, model string, stream bool) (in
 	}
 	// TODO: 非流模式下不计算图片token数量
 	if !constant.GetMediaTokenNotStream && !stream {
-		return 1000, nil
-	}
-	// 是否统计图片token
-	if !constant.GetMediaToken {
-		return 1000, nil
+		return 256, nil
 	}
 	// 同步One API的图片计费逻辑
 	if imageUrl.Detail == "auto" || imageUrl.Detail == "" {
@@ -108,6 +102,13 @@ func getImageToken(imageUrl *dto.MessageImageUrl, model string, stream bool) (in
 		tileTokens = 5667
 		baseTokens = 2833
 	}
+	// 是否统计图片token
+	if !constant.GetMediaToken {
+		return 3 * baseTokens, nil
+	}
+	if info.ChannelType == common.ChannelTypeGemini || info.ChannelType == common.ChannelTypeVertexAi || info.ChannelType == common.ChannelTypeAnthropic {
+		return 3 * baseTokens, nil
+	}
 	var config image.Config
 	var err error
 	var format string
@@ -157,9 +158,9 @@ func getImageToken(imageUrl *dto.MessageImageUrl, model string, stream bool) (in
 	return tiles*tileTokens + baseTokens, nil
 }
 
-func CountTokenChatRequest(request dto.GeneralOpenAIRequest, model string) (int, error) {
+func CountTokenChatRequest(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) (int, error) {
 	tkm := 0
-	msgTokens, err := CountTokenMessages(request.Messages, model, request.Stream)
+	msgTokens, err := CountTokenMessages(info, request.Messages, request.Model, request.Stream)
 	if err != nil {
 		return 0, err
 	}
@@ -181,7 +182,7 @@ func CountTokenChatRequest(request dto.GeneralOpenAIRequest, model string) (int,
 				countStr += fmt.Sprintf("%v", tool.Function.Parameters)
 			}
 		}
-		toolTokens, err := CountTokenInput(countStr, model)
+		toolTokens, err := CountTokenInput(countStr, request.Model)
 		if err != nil {
 			return 0, err
 		}
@@ -258,7 +259,7 @@ func CountTokenRealtime(info *relaycommon.RelayInfo, request dto.RealtimeEvent,
 	return textToken, audioToken, nil
 }
 
-func CountTokenMessages(messages []dto.Message, model string, stream bool) (int, error) {
+func CountTokenMessages(info *relaycommon.RelayInfo, messages []dto.Message, model string, stream bool) (int, error) {
 	//recover when panic
 	tokenEncoder := getTokenEncoder(model)
 	// Reference:
@@ -292,7 +293,7 @@ func CountTokenMessages(messages []dto.Message, model string, stream bool) (int,
 				for _, m := range arrayContent {
 					if m.Type == dto.ContentTypeImageURL {
 						imageUrl := m.ImageUrl.(dto.MessageImageUrl)
-						imageTokenNum, err := getImageToken(&imageUrl, model, stream)
+						imageTokenNum, err := getImageToken(info, &imageUrl, model, stream)
 						if err != nil {
 							return 0, err
 						}

setting/group_ratio.go

@@ -1,33 +1,47 @@
-package common
+package setting
 
 import (
 	"encoding/json"
 	"errors"
+	"one-api/common"
 )
 
-var GroupRatio = map[string]float64{
+var groupRatio = map[string]float64{
 	"default": 1,
 	"vip":     1,
 	"svip":    1,
 }
 
+func GetGroupRatioCopy() map[string]float64 {
+	groupRatioCopy := make(map[string]float64)
+	for k, v := range groupRatio {
+		groupRatioCopy[k] = v
+	}
+	return groupRatioCopy
+}
+
+func ContainsGroupRatio(name string) bool {
+	_, ok := groupRatio[name]
+	return ok
+}
+
 func GroupRatio2JSONString() string {
-	jsonBytes, err := json.Marshal(GroupRatio)
+	jsonBytes, err := json.Marshal(groupRatio)
 	if err != nil {
-		SysError("error marshalling model ratio: " + err.Error())
+		common.SysError("error marshalling model ratio: " + err.Error())
 	}
 	return string(jsonBytes)
 }
 
 func UpdateGroupRatioByJSONString(jsonStr string) error {
-	GroupRatio = make(map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &GroupRatio)
+	groupRatio = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &groupRatio)
 }
 
 func GetGroupRatio(name string) float64 {
-	ratio, ok := GroupRatio[name]
+	ratio, ok := groupRatio[name]
 	if !ok {
-		SysError("group ratio not found: " + name)
+		common.SysError("group ratio not found: " + name)
 		return 1
 	}
 	return ratio

setting/operation_setting.go

@@ -0,0 +1,3 @@
+package setting
+
+var DemoSiteEnabled = false

setting/user_usable_group.go

@@ -0,0 +1,52 @@
+package setting
+
+import (
+	"encoding/json"
+	"one-api/common"
+)
+
+var userUsableGroups = map[string]string{
+	"default": "默认分组",
+	"vip":     "vip分组",
+}
+
+func GetUserUsableGroupsCopy() map[string]string {
+	copyUserUsableGroups := make(map[string]string)
+	for k, v := range userUsableGroups {
+		copyUserUsableGroups[k] = v
+	}
+	return copyUserUsableGroups
+}
+
+func UserUsableGroups2JSONString() string {
+	jsonBytes, err := json.Marshal(userUsableGroups)
+	if err != nil {
+		common.SysError("error marshalling user groups: " + err.Error())
+	}
+	return string(jsonBytes)
+}
+
+func UpdateUserUsableGroupsByJSONString(jsonStr string) error {
+	userUsableGroups = make(map[string]string)
+	return json.Unmarshal([]byte(jsonStr), &userUsableGroups)
+}
+
+func GetUserUsableGroups(userGroup string) map[string]string {
+	groupsCopy := GetUserUsableGroupsCopy()
+	if userGroup == "" {
+		if _, ok := groupsCopy["default"]; !ok {
+			groupsCopy["default"] = "default"
+		}
+	}
+	// 如果userGroup不在UserUsableGroups中，返回UserUsableGroups + userGroup
+	if _, ok := groupsCopy[userGroup]; !ok {
+		groupsCopy[userGroup] = "用户分组"
+	}
+	// 如果userGroup在UserUsableGroups中，返回UserUsableGroups
+	return groupsCopy
+}
+
+func GroupInUserUsableGroups(groupName string) bool {
+	_, ok := userUsableGroups[groupName]
+	return ok
+}

web/index.html

@@ -1,5 +1,5 @@
 <!doctype html>
-<html lang="en">
+<html lang="zh">
   <head>
     <meta charset="utf-8" />
     <link rel="icon" href="/logo.png" />