feat: able to set multiple subnets

* 修正百度模型名称

* 更新百度模型名称，并保留旧版兼容以及修正单价

* chore: add more model and adjust order

---------

Co-authored-by: JustSong <songquanpeng@foxmail.com>

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

@@ -20,6 +20,13 @@ jobs:
       - name: Check out the repo
         uses: actions/checkout@v3
 
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi      
+
       - name: Save version info
         run: |
           git describe --tags > VERSION 

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

@@ -20,6 +20,13 @@ jobs:
       - name: Check out the repo
         uses: actions/checkout@v3
 
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi        
+
       - name: Save version info
         run: |
           git describe --tags > VERSION 

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

@@ -21,6 +21,13 @@ jobs:
       - name: Check out the repo
         uses: actions/checkout@v3
 
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi
+
       - name: Save version info
         run: |
           git describe --tags > VERSION 

.github/workflows/linux-release.yml

@@ -20,6 +20,12 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 0
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi
       - uses: actions/setup-node@v3
         with:
           node-version: 16

.github/workflows/macos-release.yml

@@ -20,6 +20,12 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 0
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi
       - uses: actions/setup-node@v3
         with:
           node-version: 16

.github/workflows/windows-release.yml

@@ -23,6 +23,12 @@ jobs:
         uses: actions/checkout@v3
         with:
           fetch-depth: 0
+      - name: Check repository URL
+        run: |
+          REPO_URL=$(git config --get remote.origin.url)
+          if [[ $REPO_URL == *"pro" ]]; then
+            exit 1
+          fi
       - uses: actions/setup-node@v3
         with:
           node-version: 16

.gitignore

@@ -8,3 +8,4 @@ build
 logs
 data
 /web/node_modules
+cmd.md

Dockerfile

@@ -12,6 +12,10 @@ WORKDIR /web/berry
 RUN npm install
 RUN DISABLE_ESLINT_PLUGIN='true' REACT_APP_VERSION=$(cat VERSION) npm run build
 
+WORKDIR /web/air
+RUN npm install
+RUN DISABLE_ESLINT_PLUGIN='true' REACT_APP_VERSION=$(cat VERSION) npm run build
+
 FROM golang AS builder2
 
 ENV GO111MODULE=on \

README.en.md

@@ -241,17 +241,19 @@ If the channel ID is not provided, load balancing will be used to distribute the
     + Example: `SESSION_SECRET=random_string`
 3. `SQL_DSN`: When set, the specified database will be used instead of SQLite. Please use MySQL version 8.0.
     + Example: `SQL_DSN=root:123456@tcp(localhost:3306)/oneapi`
-4. `FRONTEND_BASE_URL`: When set, the specified frontend address will be used instead of the backend address.
+4. `LOG_SQL_DSN`: When set, a separate database will be used for the `logs` table; please use MySQL or PostgreSQL.
+    + Example: `LOG_SQL_DSN=root:123456@tcp(localhost:3306)/oneapi-logs`
+5. `FRONTEND_BASE_URL`: When set, the specified frontend address will be used instead of the backend address.
     + Example: `FRONTEND_BASE_URL=https://openai.justsong.cn`
-5. `SYNC_FREQUENCY`: When set, the system will periodically sync configurations from the database, with the unit in seconds. If not set, no sync will happen.
+6. `SYNC_FREQUENCY`: When set, the system will periodically sync configurations from the database, with the unit in seconds. If not set, no sync will happen.
     + Example: `SYNC_FREQUENCY=60`
-6. `NODE_TYPE`: When set, specifies the node type. Valid values are `master` and `slave`. If not set, it defaults to `master`.
+7. `NODE_TYPE`: When set, specifies the node type. Valid values are `master` and `slave`. If not set, it defaults to `master`.
     + Example: `NODE_TYPE=slave`
-7. `CHANNEL_UPDATE_FREQUENCY`: When set, it periodically updates the channel balances, with the unit in minutes. If not set, no update will happen.
+8. `CHANNEL_UPDATE_FREQUENCY`: When set, it periodically updates the channel balances, with the unit in minutes. If not set, no update will happen.
     + Example: `CHANNEL_UPDATE_FREQUENCY=1440`
-8. `CHANNEL_TEST_FREQUENCY`: When set, it periodically tests the channels, with the unit in minutes. If not set, no test will happen.
+9. `CHANNEL_TEST_FREQUENCY`: When set, it periodically tests the channels, with the unit in minutes. If not set, no test will happen.
     + Example: `CHANNEL_TEST_FREQUENCY=1440`
-9. `POLLING_INTERVAL`: The time interval (in seconds) between requests when updating channel balances and testing channel availability. Default is no interval.
+10. `POLLING_INTERVAL`: The time interval (in seconds) between requests when updating channel balances and testing channel availability. Default is no interval.
     + Example: `POLLING_INTERVAL=5`
 
 ### Command Line Parameters

README.ja.md

@@ -242,17 +242,18 @@ graph LR
     + 例: `SESSION_SECRET=random_string`
 3. `SQL_DSN`: 設定すると、SQLite の代わりに指定したデータベースが使用されます。MySQL バージョン 8.0 を使用してください。
     + 例: `SQL_DSN=root:123456@tcp(localhost:3306)/oneapi`
-4. `FRONTEND_BASE_URL`: 設定されると、バックエンドアドレスではなく、指定されたフロントエンドアドレスが使われる。
+4. `LOG_SQL_DSN`: を設定すると、`logs`テーブルには独立したデータベースが使用されます。MySQLまたはPostgreSQLを使用してください。
+5. `FRONTEND_BASE_URL`: 設定されると、バックエンドアドレスではなく、指定されたフロントエンドアドレスが使われる。
     + 例: `FRONTEND_BASE_URL=https://openai.justsong.cn`
-5. `SYNC_FREQUENCY`: 設定された場合、システムは定期的にデータベースからコンフィグを秒単位で同期する。設定されていない場合、同期は行われません。
+6. `SYNC_FREQUENCY`: 設定された場合、システムは定期的にデータベースからコンフィグを秒単位で同期する。設定されていない場合、同期は行われません。
     + 例: `SYNC_FREQUENCY=60`
-6. `NODE_TYPE`: 設定すると、ノードのタイプを指定する。有効な値は `master` と `slave` である。設定されていない場合、デフォルトは `master`。
+7. `NODE_TYPE`: 設定すると、ノードのタイプを指定する。有効な値は `master` と `slave` である。設定されていない場合、デフォルトは `master`。
     + 例: `NODE_TYPE=slave`
-7. `CHANNEL_UPDATE_FREQUENCY`: 設定すると、チャンネル残高を分単位で定期的に更新する。設定されていない場合、更新は行われません。
+8. `CHANNEL_UPDATE_FREQUENCY`: 設定すると、チャンネル残高を分単位で定期的に更新する。設定されていない場合、更新は行われません。
     + 例: `CHANNEL_UPDATE_FREQUENCY=1440`
-8. `CHANNEL_TEST_FREQUENCY`: 設定すると、チャンネルを定期的にテストする。設定されていない場合、テストは行われません。
+9. `CHANNEL_TEST_FREQUENCY`: 設定すると、チャンネルを定期的にテストする。設定されていない場合、テストは行われません。
     + 例: `CHANNEL_TEST_FREQUENCY=1440`
-9. `POLLING_INTERVAL`: チャネル残高の更新とチャネルの可用性をテストするときのリクエスト間の時間間隔 (秒)。デフォルトは間隔なし。
+10. `POLLING_INTERVAL`: チャネル残高の更新とチャネルの可用性をテストするときのリクエスト間の時間間隔 (秒)。デフォルトは間隔なし。
     + 例: `POLLING_INTERVAL=5`
 
 ### コマンドラインパラメータ

README.md

@@ -78,13 +78,17 @@ _✨ 通过标准的 OpenAI API 格式访问所有的大模型，开箱即用
    + [x] [百川大模型](https://platform.baichuan-ai.com)
    + [ ] [字节云雀大模型](https://www.volcengine.com/product/ark) (WIP)
    + [x] [MINIMAX](https://api.minimax.chat/)
+   + [x] [Groq](https://wow.groq.com/)
+   + [x] [Ollama](https://github.com/ollama/ollama)
+   + [x] [零一万物](https://platform.lingyiwanwu.com/)
+   + [x] [阶跃星辰](https://platform.stepfun.com/)
 2. 支持配置镜像以及众多[第三方代理服务](https://iamazing.cn/page/openai-api-third-party-services)。
 3. 支持通过**负载均衡**的方式访问多个渠道。
 4. 支持 **stream 模式**，可以通过流式传输实现打字机效果。
 5. 支持**多机部署**，[详见此处](#多机部署)。
-6. 支持**令牌管理**，设置令牌的过期时间和额度。
+6. 支持**令牌管理**，设置令牌的过期时间、额度、允许的 IP 范围以及允许的模型访问。
 7. 支持**兑换码管理**，支持批量生成和导出兑换码，可使用兑换码为账户进行充值。
-8. 支持**通道管理**，批量创建通道。
+8. 支持**渠道管理**，批量创建渠道。
 9. 支持**用户分组**以及**渠道分组**，支持为不同分组设置不同的倍率。
 10. 支持渠道**设置模型列表**。
 11. 支持**查看额度明细**。
@@ -98,13 +102,15 @@ _✨ 通过标准的 OpenAI API 格式访问所有的大模型，开箱即用
 19. 支持丰富的**自定义**设置，
     1. 支持自定义系统名称，logo 以及页脚。
     2. 支持自定义首页和关于页面，可以选择使用 HTML & Markdown 代码进行自定义，或者使用一个单独的网页通过 iframe 嵌入。
-20. 支持通过系统访问令牌访问管理 API（bearer token，用以替代 cookie，你可以自行抓包来查看 API 的用法）。
+20. 支持通过系统访问令牌调用管理 API，进而**在无需二开的情况下扩展和自定义** One API 的功能，详情请参考此处 [API 文档](./docs/API.md)。。
 21. 支持 Cloudflare Turnstile 用户校验。
 22. 支持用户管理，支持**多种用户登录注册方式**：
     + 邮箱登录注册（支持注册邮箱白名单）以及通过邮箱进行密码重置。
+    + 支持使用飞书进行授权登录。
     + [GitHub 开放授权](https://github.com/settings/applications/new)。
     + 微信公众号授权（需要额外部署 [WeChat Server](https://github.com/songquanpeng/wechat-server)）。
 23. 支持主题切换，设置环境变量 `THEME` 即可，默认为 `default`，欢迎 PR 更多主题，具体参考[此处](./web/README.md)。
+24. 配合 [Message Pusher](https://github.com/songquanpeng/message-pusher) 可将报警信息推送到多种 App 上。
 
 ## 部署
 ### 基于 Docker 进行部署
@@ -345,35 +351,40 @@ graph LR
      + `SQL_MAX_OPEN_CONNS`：最大打开连接数，默认为 `1000`。
        + 如果报错 `Error 1040: Too many connections`，请适当减小该值。
      + `SQL_CONN_MAX_LIFETIME`：连接的最大生命周期，默认为 `60`，单位分钟。
-4. `FRONTEND_BASE_URL`：设置之后将重定向页面请求到指定的地址，仅限从服务器设置。
+4. `LOG_SQL_DSN`：设置之后将为 `logs` 表使用独立的数据库，请使用 MySQL 或 PostgreSQL。
+5. `FRONTEND_BASE_URL`：设置之后将重定向页面请求到指定的地址，仅限从服务器设置。
    + 例子：`FRONTEND_BASE_URL=https://openai.justsong.cn`
-5. `MEMORY_CACHE_ENABLED`：启用内存缓存，会导致用户额度的更新存在一定的延迟，可选值为 `true` 和 `false`，未设置则默认为 `false`。
+6. `MEMORY_CACHE_ENABLED`：启用内存缓存，会导致用户额度的更新存在一定的延迟，可选值为 `true` 和 `false`，未设置则默认为 `false`。
    + 例子：`MEMORY_CACHE_ENABLED=true`
-6. `SYNC_FREQUENCY`：在启用缓存的情况下与数据库同步配置的频率，单位为秒，默认为 `600` 秒。
+7. `SYNC_FREQUENCY`：在启用缓存的情况下与数据库同步配置的频率，单位为秒，默认为 `600` 秒。
    + 例子：`SYNC_FREQUENCY=60`
-7. `NODE_TYPE`：设置之后将指定节点类型，可选值为 `master` 和 `slave`，未设置则默认为 `master`。
+8. `NODE_TYPE`：设置之后将指定节点类型，可选值为 `master` 和 `slave`，未设置则默认为 `master`。
    + 例子：`NODE_TYPE=slave`
-8. `CHANNEL_UPDATE_FREQUENCY`：设置之后将定期更新渠道余额，单位为分钟，未设置则不进行更新。
+9. `CHANNEL_UPDATE_FREQUENCY`：设置之后将定期更新渠道余额，单位为分钟，未设置则不进行更新。
    + 例子：`CHANNEL_UPDATE_FREQUENCY=1440`
-9. `CHANNEL_TEST_FREQUENCY`：设置之后将定期检查渠道，单位为分钟，未设置则不进行检查。
+10. `CHANNEL_TEST_FREQUENCY`：设置之后将定期检查渠道，单位为分钟，未设置则不进行检查。
    + 例子：`CHANNEL_TEST_FREQUENCY=1440`
-10. `POLLING_INTERVAL`：批量更新渠道余额以及测试可用性时的请求间隔，单位为秒，默认无间隔。
+11. `POLLING_INTERVAL`：批量更新渠道余额以及测试可用性时的请求间隔，单位为秒，默认无间隔。
     + 例子：`POLLING_INTERVAL=5`
-11. `BATCH_UPDATE_ENABLED`：启用数据库批量更新聚合，会导致用户额度的更新存在一定的延迟可选值为 `true` 和 `false`，未设置则默认为 `false`。
+12. `BATCH_UPDATE_ENABLED`：启用数据库批量更新聚合，会导致用户额度的更新存在一定的延迟可选值为 `true` 和 `false`，未设置则默认为 `false`。
     + 例子：`BATCH_UPDATE_ENABLED=true`
     + 如果你遇到了数据库连接数过多的问题，可以尝试启用该选项。
-12. `BATCH_UPDATE_INTERVAL=5`：批量更新聚合的时间间隔，单位为秒，默认为 `5`。
+13. `BATCH_UPDATE_INTERVAL=5`：批量更新聚合的时间间隔，单位为秒，默认为 `5`。
     + 例子：`BATCH_UPDATE_INTERVAL=5`
-13. 请求频率限制：
+14. 请求频率限制：
     + `GLOBAL_API_RATE_LIMIT`：全局 API 速率限制（除中继请求外），单 ip 三分钟内的最大请求数，默认为 `180`。
     + `GLOBAL_WEB_RATE_LIMIT`：全局 Web 速率限制，单 ip 三分钟内的最大请求数，默认为 `60`。
-14. 编码器缓存设置：
+15. 编码器缓存设置：
     + `TIKTOKEN_CACHE_DIR`：默认程序启动时会联网下载一些通用的词元的编码，如：`gpt-3.5-turbo`，在一些网络环境不稳定，或者离线情况，可能会导致启动有问题，可以配置此目录缓存数据，可迁移到离线环境。
     + `DATA_GYM_CACHE_DIR`：目前该配置作用与 `TIKTOKEN_CACHE_DIR` 一致，但是优先级没有它高。
-15. `RELAY_TIMEOUT`：中继超时设置，单位为秒，默认不设置超时时间。
-16. `SQLITE_BUSY_TIMEOUT`：SQLite 锁等待超时设置，单位为毫秒，默认 `3000`。
-17. `GEMINI_SAFETY_SETTING`：Gemini 的安全设置，默认 `BLOCK_NONE`。
-18. `THEME`：系统的主题设置，默认为 `default`，具体可选值参考[此处](./web/README.md)。
+16. `RELAY_TIMEOUT`：中继超时设置，单位为秒，默认不设置超时时间。
+17. `SQLITE_BUSY_TIMEOUT`：SQLite 锁等待超时设置，单位为毫秒，默认 `3000`。
+18. `GEMINI_SAFETY_SETTING`：Gemini 的安全设置，默认 `BLOCK_NONE`。
+19. `THEME`：系统的主题设置，默认为 `default`，具体可选值参考[此处](./web/README.md)。
+20. `ENABLE_METRIC`：是否根据请求成功率禁用渠道，默认不开启，可选值为 `true` 和 `false`。
+21. `METRIC_QUEUE_SIZE`：请求成功率统计队列大小，默认为 `10`。
+22. `METRIC_SUCCESS_RATE_THRESHOLD`：请求成功率阈值，默认为 `0.8`。
+23. `INITIAL_ROOT_TOKEN`：如果设置了该值，则在系统首次启动时会自动创建一个值为该环境变量值的 root 用户令牌。
 
 ### 命令行参数
 1. `--port <port_number>`: 指定服务器监听的端口号，默认为 `3000`。
@@ -412,7 +423,7 @@ https://openai.justsong.cn
    + 检查你的接口地址和 API Key 有没有填对。
    + 检查是否启用了 HTTPS，浏览器会拦截 HTTPS 域名下的 HTTP 请求。
 6. 报错：`当前分组负载已饱和，请稍后再试`
-   + 上游通道 429 了。
+   + 上游渠道 429 了。
 7. 升级之后我的数据会丢失吗？
    + 如果使用 MySQL，不会。
    + 如果使用 SQLite，需要按照我所给的部署命令挂载 volume 持久化 one-api.db 数据库文件，否则容器重启后数据会丢失。
@@ -420,8 +431,8 @@ https://openai.justsong.cn
    + 一般情况下不需要，系统将在初始化的时候自动调整。
    + 如果需要的话，我会在更新日志中说明，并给出脚本。
 9. 手动修改数据库后报错：`数据库一致性已被破坏，请联系管理员`？
-   + 这是检测到 ability 表里有些记录的通道 id 是不存在的，这大概率是因为你删了 channel 表里的记录但是没有同步在 ability 表里清理无效的通道。
-   + 对于每一个通道，其所支持的模型都需要有一个专门的 ability 表的记录，表示该通道支持该模型。
+   + 这是检测到 ability 表里有些记录的渠道 id 是不存在的，这大概率是因为你删了 channel 表里的记录但是没有同步在 ability 表里清理无效的渠道。
+   + 对于每一个渠道，其所支持的模型都需要有一个专门的 ability 表的记录，表示该渠道支持该模型。
 
 ## 相关项目
 * [FastGPT](https://github.com/labring/FastGPT): 基于 LLM 大语言模型的知识库问答系统

common/blacklist/main.go

@@ -0,0 +1,29 @@
+package blacklist
+
+import (
+	"fmt"
+	"sync"
+)
+
+var blackList sync.Map
+
+func init() {
+	blackList = sync.Map{}
+}
+
+func userId2Key(id int) string {
+	return fmt.Sprintf("userid_%d", id)
+}
+
+func BanUser(id int) {
+	blackList.Store(userId2Key(id), true)
+}
+
+func UnbanUser(id int) {
+	blackList.Delete(userId2Key(id))
+}
+
+func IsUserBanned(id int) bool {
+	_, ok := blackList.Load(userId2Key(id))
+	return ok
+}

common/config/config.go

@@ -1,7 +1,7 @@
 package config
 
 import (
-	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/env"
 	"os"
 	"strconv"
 	"sync"
@@ -52,6 +52,7 @@ var EmailDomainWhitelist = []string{
 }
 
 var DebugEnabled = os.Getenv("DEBUG") == "true"
+var DebugSQLEnabled = os.Getenv("DEBUG_SQL") == "true"
 var MemoryCacheEnabled = os.Getenv("MEMORY_CACHE_ENABLED") == "true"
 
 var LogConsumeEnabled = true
@@ -65,21 +66,27 @@ var SMTPToken = ""
 var GitHubClientId = ""
 var GitHubClientSecret = ""
 
+var LarkClientId = ""
+var LarkClientSecret = ""
+
 var WeChatServerAddress = ""
 var WeChatServerToken = ""
 var WeChatAccountQRCodeImageURL = ""
 
+var MessagePusherAddress = ""
+var MessagePusherToken = ""
+
 var TurnstileSiteKey = ""
 var TurnstileSecretKey = ""
 
-var QuotaForNewUser = 0
-var QuotaForInviter = 0
-var QuotaForInvitee = 0
+var QuotaForNewUser int64 = 0
+var QuotaForInviter int64 = 0
+var QuotaForInvitee int64 = 0
 var ChannelDisableThreshold = 5.0
 var AutomaticDisableChannelEnabled = false
 var AutomaticEnableChannelEnabled = false
-var QuotaRemindThreshold = 1000
-var PreConsumedQuota = 500
+var QuotaRemindThreshold int64 = 1000
+var PreConsumedQuota int64 = 500
 var ApproximateTokenEnabled = false
 var RetryTimes = 0
 
@@ -90,28 +97,29 @@ var IsMasterNode = os.Getenv("NODE_TYPE") != "slave"
 var requestInterval, _ = strconv.Atoi(os.Getenv("POLLING_INTERVAL"))
 var RequestInterval = time.Duration(requestInterval) * time.Second
 
-var SyncFrequency = helper.GetOrDefaultEnvInt("SYNC_FREQUENCY", 10*60) // unit is second
+var SyncFrequency = env.Int("SYNC_FREQUENCY", 10*60) // unit is second
 
 var BatchUpdateEnabled = false
-var BatchUpdateInterval = helper.GetOrDefaultEnvInt("BATCH_UPDATE_INTERVAL", 5)
+var BatchUpdateInterval = env.Int("BATCH_UPDATE_INTERVAL", 5)
 
-var RelayTimeout = helper.GetOrDefaultEnvInt("RELAY_TIMEOUT", 0) // unit is second
+var RelayTimeout = env.Int("RELAY_TIMEOUT", 0) // unit is second
 
-var GeminiSafetySetting = helper.GetOrDefaultEnvString("GEMINI_SAFETY_SETTING", "BLOCK_NONE")
+var GeminiSafetySetting = env.String("GEMINI_SAFETY_SETTING", "BLOCK_NONE")
 
-var Theme = helper.GetOrDefaultEnvString("THEME", "default")
+var Theme = env.String("THEME", "default")
 var ValidThemes = map[string]bool{
 	"default": true,
 	"berry":   true,
+	"air":     true,
 }
 
 // All duration's unit is seconds
 // Shouldn't larger then RateLimitKeyExpirationDuration
 var (
-	GlobalApiRateLimitNum            = helper.GetOrDefaultEnvInt("GLOBAL_API_RATE_LIMIT", 180)
+	GlobalApiRateLimitNum            = env.Int("GLOBAL_API_RATE_LIMIT", 180)
 	GlobalApiRateLimitDuration int64 = 3 * 60
 
-	GlobalWebRateLimitNum            = helper.GetOrDefaultEnvInt("GLOBAL_WEB_RATE_LIMIT", 60)
+	GlobalWebRateLimitNum            = env.Int("GLOBAL_WEB_RATE_LIMIT", 60)
 	GlobalWebRateLimitDuration int64 = 3 * 60
 
 	UploadRateLimitNum            = 10
@@ -125,3 +133,11 @@ var (
 )
 
 var RateLimitKeyExpirationDuration = 20 * time.Minute
+
+var EnableMetric = env.Bool("ENABLE_METRIC", false)
+var MetricQueueSize = env.Int("METRIC_QUEUE_SIZE", 10)
+var MetricSuccessRateThreshold = env.Float64("METRIC_SUCCESS_RATE_THRESHOLD", 0.8)
+var MetricSuccessChanSize = env.Int("METRIC_SUCCESS_CHAN_SIZE", 1024)
+var MetricFailChanSize = env.Int("METRIC_FAIL_CHAN_SIZE", 128)
+
+var InitialRootToken = os.Getenv("INITIAL_ROOT_TOKEN")

common/constants.go

@@ -15,6 +15,7 @@ const (
 const (
 	UserStatusEnabled  = 1 // don't use 0, 0 is the default value!
 	UserStatusDisabled = 2 // also don't use 0
+	UserStatusDeleted  = 3
 )
 
 const (
@@ -38,35 +39,41 @@ const (
 )
 
 const (
-	ChannelTypeUnknown        = 0
-	ChannelTypeOpenAI         = 1
-	ChannelTypeAPI2D          = 2
-	ChannelTypeAzure          = 3
-	ChannelTypeCloseAI        = 4
-	ChannelTypeOpenAISB       = 5
-	ChannelTypeOpenAIMax      = 6
-	ChannelTypeOhMyGPT        = 7
-	ChannelTypeCustom         = 8
-	ChannelTypeAILS           = 9
-	ChannelTypeAIProxy        = 10
-	ChannelTypePaLM           = 11
-	ChannelTypeAPI2GPT        = 12
-	ChannelTypeAIGC2D         = 13
-	ChannelTypeAnthropic      = 14
-	ChannelTypeBaidu          = 15
-	ChannelTypeZhipu          = 16
-	ChannelTypeAli            = 17
-	ChannelTypeXunfei         = 18
-	ChannelType360            = 19
-	ChannelTypeOpenRouter     = 20
-	ChannelTypeAIProxyLibrary = 21
-	ChannelTypeFastGPT        = 22
-	ChannelTypeTencent        = 23
-	ChannelTypeGemini         = 24
-	ChannelTypeMoonshot       = 25
-	ChannelTypeBaichuan       = 26
-	ChannelTypeMinimax        = 27
-	ChannelTypeMistral        = 28
+	ChannelTypeUnknown = iota
+	ChannelTypeOpenAI
+	ChannelTypeAPI2D
+	ChannelTypeAzure
+	ChannelTypeCloseAI
+	ChannelTypeOpenAISB
+	ChannelTypeOpenAIMax
+	ChannelTypeOhMyGPT
+	ChannelTypeCustom
+	ChannelTypeAILS
+	ChannelTypeAIProxy
+	ChannelTypePaLM
+	ChannelTypeAPI2GPT
+	ChannelTypeAIGC2D
+	ChannelTypeAnthropic
+	ChannelTypeBaidu
+	ChannelTypeZhipu
+	ChannelTypeAli
+	ChannelTypeXunfei
+	ChannelType360
+	ChannelTypeOpenRouter
+	ChannelTypeAIProxyLibrary
+	ChannelTypeFastGPT
+	ChannelTypeTencent
+	ChannelTypeGemini
+	ChannelTypeMoonshot
+	ChannelTypeBaichuan
+	ChannelTypeMinimax
+	ChannelTypeMistral
+	ChannelTypeGroq
+	ChannelTypeOllama
+	ChannelTypeLingYiWanWu
+	ChannelTypeStepFun
+
+	ChannelTypeDummy
 )
 
 var ChannelBaseURLs = []string{
@@ -99,6 +106,10 @@ var ChannelBaseURLs = []string{
 	"https://api.baichuan-ai.com",               // 26
 	"https://api.minimax.chat",                  // 27
 	"https://api.mistral.ai",                    // 28
+	"https://api.groq.com/openai",               // 29
+	"http://localhost:11434",                    // 30
+	"https://api.lingyiwanwu.com",               // 31
+	"https://api.stepfun.com",                   // 32
 }
 
 const (

common/conv/any.go

@@ -0,0 +1,6 @@
+package conv
+
+func AsString(v any) string {
+	str, _ := v.(string)
+	return str
+}

common/database.go

@@ -1,9 +1,12 @@
 package common
 
-import "github.com/songquanpeng/one-api/common/helper"
+import (
+	"github.com/songquanpeng/one-api/common/env"
+)
 
 var UsingSQLite = false
 var UsingPostgreSQL = false
+var UsingMySQL = false
 
 var SQLitePath = "one-api.db"
-var SQLiteBusyTimeout = helper.GetOrDefaultEnvInt("SQLITE_BUSY_TIMEOUT", 3000)
+var SQLiteBusyTimeout = env.Int("SQLITE_BUSY_TIMEOUT", 3000)

common/env/helper.go

@@ -0,0 +1,42 @@
+package env
+
+import (
+	"os"
+	"strconv"
+)
+
+func Bool(env string, defaultValue bool) bool {
+	if env == "" || os.Getenv(env) == "" {
+		return defaultValue
+	}
+	return os.Getenv(env) == "true"
+}
+
+func Int(env string, defaultValue int) int {
+	if env == "" || os.Getenv(env) == "" {
+		return defaultValue
+	}
+	num, err := strconv.Atoi(os.Getenv(env))
+	if err != nil {
+		return defaultValue
+	}
+	return num
+}
+
+func Float64(env string, defaultValue float64) float64 {
+	if env == "" || os.Getenv(env) == "" {
+		return defaultValue
+	}
+	num, err := strconv.ParseFloat(os.Getenv(env), 64)
+	if err != nil {
+		return defaultValue
+	}
+	return num
+}
+
+func String(env string, defaultValue string) string {
+	if env == "" || os.Getenv(env) == "" {
+		return defaultValue
+	}
+	return os.Getenv(env)
+}

common/helper/helper.go

@@ -3,12 +3,10 @@ package helper
 import (
 	"fmt"
 	"github.com/google/uuid"
-	"github.com/songquanpeng/one-api/common/logger"
 	"html/template"
 	"log"
 	"math/rand"
 	"net"
-	"os"
 	"os/exec"
 	"runtime"
 	"strconv"
@@ -187,6 +185,10 @@ func GetTimeString() string {
 	return fmt.Sprintf("%s%d", now.Format("20060102150405"), now.UnixNano()%1e9)
 }
 
+func GenRequestID() string {
+	return GetTimeString() + GetRandomNumberString(8)
+}
+
 func Max(a int, b int) int {
 	if a >= b {
 		return a
@@ -195,25 +197,6 @@ func Max(a int, b int) int {
 	}
 }
 
-func GetOrDefaultEnvInt(env string, defaultValue int) int {
-	if env == "" || os.Getenv(env) == "" {
-		return defaultValue
-	}
-	num, err := strconv.Atoi(os.Getenv(env))
-	if err != nil {
-		logger.SysError(fmt.Sprintf("failed to parse %s: %s, using default value: %d", env, err.Error(), defaultValue))
-		return defaultValue
-	}
-	return num
-}
-
-func GetOrDefaultEnvString(env string, defaultValue string) string {
-	if env == "" || os.Getenv(env) == "" {
-		return defaultValue
-	}
-	return os.Getenv(env)
-}
-
 func AssignOrDefault(value string, defaultValue string) string {
 	if len(value) != 0 {
 		return value

common/logger/logger.go

@@ -4,6 +4,8 @@ import (
 	"context"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
 	"io"
 	"log"
 	"os"
@@ -19,9 +21,6 @@ const (
 	loggerError = "ERR"
 )
 
-const maxLogCount = 1000000
-
-var logCount int
 var setupLogLock sync.Mutex
 var setupLogWorking bool
 
@@ -57,7 +56,9 @@ func SysError(s string) {
 }
 
 func Debug(ctx context.Context, msg string) {
-	logHelper(ctx, loggerDEBUG, msg)
+	if config.DebugEnabled {
+		logHelper(ctx, loggerDEBUG, msg)
+	}
 }
 
 func Info(ctx context.Context, msg string) {
@@ -94,11 +95,12 @@ func logHelper(ctx context.Context, level string, msg string) {
 		writer = gin.DefaultWriter
 	}
 	id := ctx.Value(RequestIdKey)
+	if id == nil {
+		id = helper.GenRequestID()
+	}
 	now := time.Now()
 	_, _ = fmt.Fprintf(writer, "[%s] %v | %s | %s \n", level, now.Format("2006/01/02 - 15:04:05"), id, msg)
-	logCount++ // we don't need accurate count, so no lock here
-	if logCount > maxLogCount && !setupLogWorking {
-		logCount = 0
+	if !setupLogWorking {
 		setupLogWorking = true
 		go func() {
 			SetupLogger()

common/message/email.go

@@ -1,4 +1,4 @@
-package common
+package message
 
 import (
 	"crypto/rand"
@@ -12,6 +12,9 @@ import (
 )
 
 func SendEmail(subject string, receiver string, content string) error {
+	if receiver == "" {
+		return fmt.Errorf("receiver is empty")
+	}
 	if config.SMTPFrom == "" { // for compatibility
 		config.SMTPFrom = config.SMTPAccount
 	}

common/message/main.go

@@ -0,0 +1,22 @@
+package message
+
+import (
+	"fmt"
+	"github.com/songquanpeng/one-api/common/config"
+)
+
+const (
+	ByAll           = "all"
+	ByEmail         = "email"
+	ByMessagePusher = "message_pusher"
+)
+
+func Notify(by string, title string, description string, content string) error {
+	if by == ByEmail {
+		return SendEmail(title, config.RootUserEmail, content)
+	}
+	if by == ByMessagePusher {
+		return SendMessage(title, description, content)
+	}
+	return fmt.Errorf("unknown notify method: %s", by)
+}

common/message/message-pusher.go

@@ -0,0 +1,53 @@
+package message
+
+import (
+	"bytes"
+	"encoding/json"
+	"errors"
+	"github.com/songquanpeng/one-api/common/config"
+	"net/http"
+)
+
+type request struct {
+	Title       string `json:"title"`
+	Description string `json:"description"`
+	Content     string `json:"content"`
+	URL         string `json:"url"`
+	Channel     string `json:"channel"`
+	Token       string `json:"token"`
+}
+
+type response struct {
+	Success bool   `json:"success"`
+	Message string `json:"message"`
+}
+
+func SendMessage(title string, description string, content string) error {
+	if config.MessagePusherAddress == "" {
+		return errors.New("message pusher address is not set")
+	}
+	req := request{
+		Title:       title,
+		Description: description,
+		Content:     content,
+		Token:       config.MessagePusherToken,
+	}
+	data, err := json.Marshal(req)
+	if err != nil {
+		return err
+	}
+	resp, err := http.Post(config.MessagePusherAddress,
+		"application/json", bytes.NewBuffer(data))
+	if err != nil {
+		return err
+	}
+	var res response
+	err = json.NewDecoder(resp.Body).Decode(&res)
+	if err != nil {
+		return err
+	}
+	if !res.Success {
+		return errors.New(res.Message)
+	}
+	return nil
+}

common/model-ratio.go

@@ -4,7 +4,6 @@ import (
 	"encoding/json"
 	"github.com/songquanpeng/one-api/common/logger"
 	"strings"
-	"time"
 )
 
 const (
@@ -31,7 +30,7 @@ var ModelRatio = map[string]float64{
 	"gpt-4-0125-preview":      5,    // $0.01 / 1K tokens
 	"gpt-4-turbo-preview":     5,    // $0.01 / 1K tokens
 	"gpt-4-vision-preview":    5,    // $0.01 / 1K tokens
-	"gpt-3.5-turbo":           0.75, // $0.0015 / 1K tokens
+	"gpt-3.5-turbo":           0.25, // $0.0005 / 1K tokens
 	"gpt-3.5-turbo-0301":      0.75,
 	"gpt-3.5-turbo-0613":      0.75,
 	"gpt-3.5-turbo-16k":       1.5, // $0.003 / 1K tokens
@@ -63,25 +62,44 @@ var ModelRatio = map[string]float64{
 	"text-search-ada-doc-001": 10,
 	"text-moderation-stable":  0.1,
 	"text-moderation-latest":  0.1,
-	"dall-e-2":                8,     // $0.016 - $0.020 / image
-	"dall-e-3":                20,    // $0.040 - $0.120 / image
-	"claude-instant-1":        0.815, // $1.63 / 1M tokens
-	"claude-2":                5.51,  // $11.02 / 1M tokens
-	"claude-2.0":              5.51,  // $11.02 / 1M tokens
-	"claude-2.1":              5.51,  // $11.02 / 1M tokens
+	"dall-e-2":                8,  // $0.016 - $0.020 / image
+	"dall-e-3":                20, // $0.040 - $0.120 / image
+	// https://www.anthropic.com/api#pricing
+	"claude-instant-1.2":       0.8 / 1000 * USD,
+	"claude-2.0":               8.0 / 1000 * USD,
+	"claude-2.1":               8.0 / 1000 * USD,
+	"claude-3-haiku-20240307":  0.25 / 1000 * USD,
+	"claude-3-sonnet-20240229": 3.0 / 1000 * USD,
+	"claude-3-opus-20240229":   15.0 / 1000 * USD,
 	// https://cloud.baidu.com/doc/WENXINWORKSHOP/s/hlrk4akp7
-	"ERNIE-Bot":         0.8572,     // ￥0.012 / 1k tokens
-	"ERNIE-Bot-turbo":   0.5715,     // ￥0.008 / 1k tokens
-	"ERNIE-Bot-4":       0.12 * RMB, // ￥0.12 / 1k tokens
-	"ERNIE-Bot-8k":      0.024 * RMB,
-	"Embedding-V1":      0.1429, // ￥0.002 / 1k tokens
-	"PaLM-2":            1,
-	"gemini-pro":        1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
-	"gemini-pro-vision": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	"ERNIE-4.0-8K":       0.120 * RMB,
+	"ERNIE-Bot-8K-0922":  0.024 * RMB,
+	"ERNIE-3.5-8K":       0.012 * RMB,
+	"ERNIE-Lite-8K-0922": 0.008 * RMB,
+	"ERNIE-Speed-8K":     0.004 * RMB,
+	"ERNIE-3.5-4K-0205":  0.012 * RMB,
+	"ERNIE-3.5-8K-0205":  0.024 * RMB,
+	"ERNIE-3.5-8K-1222":  0.012 * RMB,
+	"ERNIE-Lite-8K":      0.003 * RMB,
+	"ERNIE-Speed-128K":   0.004 * RMB,
+	"ERNIE-Tiny-8K":      0.001 * RMB,
+	"BLOOMZ-7B":          0.004 * RMB,
+	"Embedding-V1":       0.002 * RMB,
+	"bge-large-zh":       0.002 * RMB,
+	"bge-large-en":       0.002 * RMB,
+	"tao-8k":             0.002 * RMB,
+	// https://ai.google.dev/pricing
+	"PaLM-2":                    1,
+	"gemini-pro":                1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	"gemini-pro-vision":         1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	"gemini-1.0-pro-vision-001": 1,
+	"gemini-1.0-pro-001":        1,
+	"gemini-1.5-pro":            1,
 	// https://open.bigmodel.cn/pricing
 	"glm-4":                     0.1 * RMB,
 	"glm-4v":                    0.1 * RMB,
 	"glm-3-turbo":               0.005 * RMB,
+	"embedding-2":               0.0005 * RMB,
 	"chatglm_turbo":             0.3572, // ￥0.005 / 1k tokens
 	"chatglm_pro":               0.7143, // ￥0.01 / 1k tokens
 	"chatglm_std":               0.3572, // ￥0.005 / 1k tokens
@@ -122,6 +140,15 @@ var ModelRatio = map[string]float64{
 	"mistral-medium-latest": 2.7 / 1000 * USD,
 	"mistral-large-latest":  8.0 / 1000 * USD,
 	"mistral-embed":         0.1 / 1000 * USD,
+	// https://wow.groq.com/
+	"llama2-70b-4096":    0.7 / 1000 * USD,
+	"llama2-7b-2048":     0.1 / 1000 * USD,
+	"mixtral-8x7b-32768": 0.27 / 1000 * USD,
+	"gemma-7b-it":        0.1 / 1000 * USD,
+	// https://platform.lingyiwanwu.com/docs#-计费单元
+	"yi-34b-chat-0205": 2.5 / 1000 * RMB,
+	"yi-34b-chat-200k": 12.0 / 1000 * RMB,
+	"yi-vl-plus":       6.0 / 1000 * RMB,
 }
 
 var CompletionRatio = map[string]float64{}
@@ -140,6 +167,26 @@ func init() {
 	}
 }
 
+func AddNewMissingRatio(oldRatio string) string {
+	newRatio := make(map[string]float64)
+	err := json.Unmarshal([]byte(oldRatio), &newRatio)
+	if err != nil {
+		logger.SysError("error unmarshalling old ratio: " + err.Error())
+		return oldRatio
+	}
+	for k, v := range DefaultModelRatio {
+		if _, ok := newRatio[k]; !ok {
+			newRatio[k] = v
+		}
+	}
+	jsonBytes, err := json.Marshal(newRatio)
+	if err != nil {
+		logger.SysError("error marshalling new ratio: " + err.Error())
+		return oldRatio
+	}
+	return string(jsonBytes)
+}
+
 func ModelRatio2JSONString() string {
 	jsonBytes, err := json.Marshal(ModelRatio)
 	if err != nil {
@@ -189,7 +236,7 @@ func GetCompletionRatio(name string) float64 {
 		return ratio
 	}
 	if strings.HasPrefix(name, "gpt-3.5") {
-		if strings.HasSuffix(name, "0125") {
+		if name == "gpt-3.5-turbo" || strings.HasSuffix(name, "0125") {
 			// https://openai.com/blog/new-embedding-models-and-api-updates
 			// Updated GPT-3.5 Turbo model and lower pricing
 			return 3
@@ -197,16 +244,7 @@ func GetCompletionRatio(name string) float64 {
 		if strings.HasSuffix(name, "1106") {
 			return 2
 		}
-		if name == "gpt-3.5-turbo" || name == "gpt-3.5-turbo-16k" {
-			// TODO: clear this after 2023-12-11
-			now := time.Now()
-			// https://platform.openai.com/docs/models/continuous-model-upgrades
-			// if after 2023-12-11, use 2
-			if now.After(time.Date(2023, 12, 11, 0, 0, 0, 0, time.UTC)) {
-				return 2
-			}
-		}
-		return 1.333333
+		return 4.0 / 3.0
 	}
 	if strings.HasPrefix(name, "gpt-4") {
 		if strings.HasSuffix(name, "preview") {
@@ -214,14 +252,21 @@ func GetCompletionRatio(name string) float64 {
 		}
 		return 2
 	}
-	if strings.HasPrefix(name, "claude-instant-1") {
-		return 3.38
+	if strings.HasPrefix(name, "claude-3") {
+		return 5
 	}
-	if strings.HasPrefix(name, "claude-2") {
-		return 2.965517
+	if strings.HasPrefix(name, "claude-") {
+		return 3
 	}
 	if strings.HasPrefix(name, "mistral-") {
 		return 3
 	}
+	if strings.HasPrefix(name, "gemini-") {
+		return 3
+	}
+	switch name {
+	case "llama2-70b-4096":
+		return 0.8 / 0.7
+	}
 	return 1
 }

common/network/ip.go

@@ -0,0 +1,52 @@
+package network
+
+import (
+	"context"
+	"fmt"
+	"github.com/songquanpeng/one-api/common/logger"
+	"net"
+	"strings"
+)
+
+func splitSubnets(subnets string) []string {
+	res := strings.Split(subnets, ",")
+	for i := 0; i < len(res); i++ {
+		res[i] = strings.TrimSpace(res[i])
+	}
+	return res
+}
+
+func isValidSubnet(subnet string) error {
+	_, _, err := net.ParseCIDR(subnet)
+	if err != nil {
+		return fmt.Errorf("failed to parse subnet: %w", err)
+	}
+	return nil
+}
+
+func isIpInSubnet(ctx context.Context, ip string, subnet string) bool {
+	_, ipNet, err := net.ParseCIDR(subnet)
+	if err != nil {
+		logger.Errorf(ctx, "failed to parse subnet: %s", err.Error())
+		return false
+	}
+	return ipNet.Contains(net.ParseIP(ip))
+}
+
+func IsValidSubnets(subnets string) error {
+	for _, subnet := range splitSubnets(subnets) {
+		if err := isValidSubnet(subnet); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func IsIpInSubnets(ctx context.Context, ip string, subnets string) bool {
+	for _, subnet := range splitSubnets(subnets) {
+		if isIpInSubnet(ctx, ip, subnet) {
+			return true
+		}
+	}
+	return false
+}

common/network/ip_test.go

@@ -0,0 +1,19 @@
+package network
+
+import (
+	"context"
+	"testing"
+
+	. "github.com/smartystreets/goconvey/convey"
+)
+
+func TestIsIpInSubnet(t *testing.T) {
+	ctx := context.Background()
+	ip1 := "192.168.0.5"
+	ip2 := "125.216.250.89"
+	subnet := "192.168.0.0/24"
+	Convey("TestIsIpInSubnet", t, func() {
+		So(isIpInSubnet(ctx, ip1, subnet), ShouldBeTrue)
+		So(isIpInSubnet(ctx, ip2, subnet), ShouldBeFalse)
+	})
+}

common/utils.go

@@ -5,7 +5,7 @@ import (
 	"github.com/songquanpeng/one-api/common/config"
 )
 
-func LogQuota(quota int) string {
+func LogQuota(quota int64) string {
 	if config.DisplayInCurrencyEnabled {
 		return fmt.Sprintf("＄%.6f 额度", float64(quota)/config.QuotaPerUnit)
 	} else {

controller/auth/github.go

@@ -1,4 +1,4 @@
-package controller
+package auth
 
 import (
 	"bytes"
@@ -11,6 +11,7 @@ import (
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/controller"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strconv"
@@ -159,7 +160,7 @@ func GitHubOAuth(c *gin.Context) {
 		})
 		return
 	}
-	setupLogin(&user, c)
+	controller.SetupLogin(&user, c)
 }
 
 func GitHubBind(c *gin.Context) {

controller/auth/lark.go

@@ -0,0 +1,201 @@
+package auth
+
+import (
+	"bytes"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"github.com/gin-contrib/sessions"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/controller"
+	"github.com/songquanpeng/one-api/model"
+	"net/http"
+	"strconv"
+	"time"
+)
+
+type LarkOAuthResponse struct {
+	AccessToken string `json:"access_token"`
+}
+
+type LarkUser struct {
+	Name   string `json:"name"`
+	OpenID string `json:"open_id"`
+}
+
+func getLarkUserInfoByCode(code string) (*LarkUser, error) {
+	if code == "" {
+		return nil, errors.New("无效的参数")
+	}
+	values := map[string]string{
+		"client_id":     config.LarkClientId,
+		"client_secret": config.LarkClientSecret,
+		"code":          code,
+		"grant_type":    "authorization_code",
+		"redirect_uri":  fmt.Sprintf("%s/oauth/lark", config.ServerAddress),
+	}
+	jsonData, err := json.Marshal(values)
+	if err != nil {
+		return nil, err
+	}
+	req, err := http.NewRequest("POST", "https://passport.feishu.cn/suite/passport/oauth/token", bytes.NewBuffer(jsonData))
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Accept", "application/json")
+	client := http.Client{
+		Timeout: 5 * time.Second,
+	}
+	res, err := client.Do(req)
+	if err != nil {
+		logger.SysLog(err.Error())
+		return nil, errors.New("无法连接至飞书服务器，请稍后重试！")
+	}
+	defer res.Body.Close()
+	var oAuthResponse LarkOAuthResponse
+	err = json.NewDecoder(res.Body).Decode(&oAuthResponse)
+	if err != nil {
+		return nil, err
+	}
+	req, err = http.NewRequest("GET", "https://passport.feishu.cn/suite/passport/oauth/userinfo", nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", oAuthResponse.AccessToken))
+	res2, err := client.Do(req)
+	if err != nil {
+		logger.SysLog(err.Error())
+		return nil, errors.New("无法连接至飞书服务器，请稍后重试！")
+	}
+	var larkUser LarkUser
+	err = json.NewDecoder(res2.Body).Decode(&larkUser)
+	if err != nil {
+		return nil, err
+	}
+	return &larkUser, nil
+}
+
+func LarkOAuth(c *gin.Context) {
+	session := sessions.Default(c)
+	state := c.Query("state")
+	if state == "" || session.Get("oauth_state") == nil || state != session.Get("oauth_state").(string) {
+		c.JSON(http.StatusForbidden, gin.H{
+			"success": false,
+			"message": "state is empty or not same",
+		})
+		return
+	}
+	username := session.Get("username")
+	if username != nil {
+		LarkBind(c)
+		return
+	}
+	code := c.Query("code")
+	larkUser, err := getLarkUserInfoByCode(code)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	user := model.User{
+		LarkId: larkUser.OpenID,
+	}
+	if model.IsLarkIdAlreadyTaken(user.LarkId) {
+		err := user.FillUserByLarkId()
+		if err != nil {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": err.Error(),
+			})
+			return
+		}
+	} else {
+		if config.RegisterEnabled {
+			user.Username = "lark_" + strconv.Itoa(model.GetMaxUserId()+1)
+			if larkUser.Name != "" {
+				user.DisplayName = larkUser.Name
+			} else {
+				user.DisplayName = "Lark User"
+			}
+			user.Role = common.RoleCommonUser
+			user.Status = common.UserStatusEnabled
+
+			if err := user.Insert(0); err != nil {
+				c.JSON(http.StatusOK, gin.H{
+					"success": false,
+					"message": err.Error(),
+				})
+				return
+			}
+		} else {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "管理员关闭了新用户注册",
+			})
+			return
+		}
+	}
+
+	if user.Status != common.UserStatusEnabled {
+		c.JSON(http.StatusOK, gin.H{
+			"message": "用户已被封禁",
+			"success": false,
+		})
+		return
+	}
+	controller.SetupLogin(&user, c)
+}
+
+func LarkBind(c *gin.Context) {
+	code := c.Query("code")
+	larkUser, err := getLarkUserInfoByCode(code)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	user := model.User{
+		LarkId: larkUser.OpenID,
+	}
+	if model.IsLarkIdAlreadyTaken(user.LarkId) {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "该飞书账户已被绑定",
+		})
+		return
+	}
+	session := sessions.Default(c)
+	id := session.Get("id")
+	// id := c.GetInt("id")  // critical bug!
+	user.Id = id.(int)
+	err = user.FillUserById()
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	user.LarkId = larkUser.OpenID
+	err = user.Update(false)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "bind",
+	})
+	return
+}

controller/auth/wechat.go

@@ -1,4 +1,4 @@
-package controller
+package auth
 
 import (
 	"encoding/json"
@@ -7,6 +7,7 @@ import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/controller"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strconv"
@@ -109,7 +110,7 @@ func WeChatAuth(c *gin.Context) {
 		})
 		return
 	}
-	setupLogin(&user, c)
+	controller.SetupLogin(&user, c)
 }
 
 func WeChatBind(c *gin.Context) {

controller/billing.go

@@ -8,8 +8,8 @@ import (
 )
 
 func GetSubscription(c *gin.Context) {
-	var remainQuota int
-	var usedQuota int
+	var remainQuota int64
+	var usedQuota int64
 	var err error
 	var token *model.Token
 	var expiredTime int64
@@ -60,7 +60,7 @@ func GetSubscription(c *gin.Context) {
 }
 
 func GetUsage(c *gin.Context) {
-	var quota int
+	var quota int64
 	var err error
 	var token *model.Token
 	if config.DisplayTokenStatEnabled {

controller/channel-billing.go

@@ -8,6 +8,7 @@ import (
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/monitor"
 	"github.com/songquanpeng/one-api/relay/util"
 	"io"
 	"net/http"
@@ -295,7 +296,7 @@ func UpdateChannelBalance(c *gin.Context) {
 }
 
 func updateAllChannelsBalance() error {
-	channels, err := model.GetAllChannels(0, 0, true)
+	channels, err := model.GetAllChannels(0, 0, "all")
 	if err != nil {
 		return err
 	}
@@ -313,7 +314,7 @@ func updateAllChannelsBalance() error {
 		} else {
 			// err is nil & balance <= 0 means quota is used up
 			if balance <= 0 {
-				disableChannel(channel.Id, channel.Name, "余额不足")
+				monitor.DisableChannel(channel.Id, channel.Name, "余额不足")
 			}
 		}
 		time.Sleep(config.RequestInterval)
@@ -322,15 +323,14 @@ func updateAllChannelsBalance() error {
 }
 
 func UpdateAllChannelsBalance(c *gin.Context) {
-	// TODO: make it async
-	err := updateAllChannelsBalance()
-	if err != nil {
-		c.JSON(http.StatusOK, gin.H{
-			"success": false,
-			"message": err.Error(),
-		})
-		return
-	}
+	//err := updateAllChannelsBalance()
+	//if err != nil {
+	//	c.JSON(http.StatusOK, gin.H{
+	//		"success": false,
+	//		"message": err.Error(),
+	//	})
+	//	return
+	//}
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",

controller/channel-test.go

@@ -8,8 +8,10 @@ import (
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/message"
 	"github.com/songquanpeng/one-api/middleware"
 	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/monitor"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/helper"
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
@@ -28,7 +30,7 @@ import (
 
 func buildTestRequest() *relaymodel.GeneralOpenAIRequest {
 	testRequest := &relaymodel.GeneralOpenAIRequest{
-		MaxTokens: 1,
+		MaxTokens: 2,
 		Stream:    false,
 		Model:     "gpt-3.5-turbo",
 	}
@@ -148,33 +150,7 @@ func TestChannel(c *gin.Context) {
 var testAllChannelsLock sync.Mutex
 var testAllChannelsRunning bool = false
 
-func notifyRootUser(subject string, content string) {
-	if config.RootUserEmail == "" {
-		config.RootUserEmail = model.GetRootUserEmail()
-	}
-	err := common.SendEmail(subject, config.RootUserEmail, content)
-	if err != nil {
-		logger.SysError(fmt.Sprintf("failed to send email: %s", err.Error()))
-	}
-}
-
-// disable & notify
-func disableChannel(channelId int, channelName string, reason string) {
-	model.UpdateChannelStatusById(channelId, common.ChannelStatusAutoDisabled)
-	subject := fmt.Sprintf("通道「%s」（#%d）已被禁用", channelName, channelId)
-	content := fmt.Sprintf("通道「%s」（#%d）已被禁用，原因：%s", channelName, channelId, reason)
-	notifyRootUser(subject, content)
-}
-
-// enable & notify
-func enableChannel(channelId int, channelName string) {
-	model.UpdateChannelStatusById(channelId, common.ChannelStatusEnabled)
-	subject := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
-	content := fmt.Sprintf("通道「%s」（#%d）已被启用", channelName, channelId)
-	notifyRootUser(subject, content)
-}
-
-func testAllChannels(notify bool) error {
+func testChannels(notify bool, scope string) error {
 	if config.RootUserEmail == "" {
 		config.RootUserEmail = model.GetRootUserEmail()
 	}
@@ -185,7 +161,7 @@ func testAllChannels(notify bool) error {
 	}
 	testAllChannelsRunning = true
 	testAllChannelsLock.Unlock()
-	channels, err := model.GetAllChannels(0, 0, true)
+	channels, err := model.GetAllChannels(0, 0, scope)
 	if err != nil {
 		return err
 	}
@@ -202,13 +178,17 @@ func testAllChannels(notify bool) error {
 			milliseconds := tok.Sub(tik).Milliseconds()
 			if isChannelEnabled && milliseconds > disableThreshold {
 				err = errors.New(fmt.Sprintf("响应时间 %.2fs 超过阈值 %.2fs", float64(milliseconds)/1000.0, float64(disableThreshold)/1000.0))
-				disableChannel(channel.Id, channel.Name, err.Error())
+				if config.AutomaticDisableChannelEnabled {
+					monitor.DisableChannel(channel.Id, channel.Name, err.Error())
+				} else {
+					_ = message.Notify(message.ByAll, fmt.Sprintf("渠道 %s （%d）测试超时", channel.Name, channel.Id), "", err.Error())
+				}
 			}
 			if isChannelEnabled && util.ShouldDisableChannel(openaiErr, -1) {
-				disableChannel(channel.Id, channel.Name, err.Error())
+				monitor.DisableChannel(channel.Id, channel.Name, err.Error())
 			}
 			if !isChannelEnabled && util.ShouldEnableChannel(err, openaiErr) {
-				enableChannel(channel.Id, channel.Name)
+				monitor.EnableChannel(channel.Id, channel.Name)
 			}
 			channel.UpdateResponseTime(milliseconds)
 			time.Sleep(config.RequestInterval)
@@ -217,7 +197,7 @@ func testAllChannels(notify bool) error {
 		testAllChannelsRunning = false
 		testAllChannelsLock.Unlock()
 		if notify {
-			err := common.SendEmail("通道测试完成", config.RootUserEmail, "通道测试完成，如果没有收到禁用通知，说明所有通道都正常")
+			err := message.Notify(message.ByAll, "渠道测试完成", "", "渠道测试完成，如果没有收到禁用通知，说明所有渠道都正常")
 			if err != nil {
 				logger.SysError(fmt.Sprintf("failed to send email: %s", err.Error()))
 			}
@@ -226,8 +206,12 @@ func testAllChannels(notify bool) error {
 	return nil
 }
 
-func TestAllChannels(c *gin.Context) {
-	err := testAllChannels(true)
+func TestChannels(c *gin.Context) {
+	scope := c.Query("scope")
+	if scope == "" {
+		scope = "all"
+	}
+	err := testChannels(true, scope)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -246,7 +230,7 @@ func AutomaticallyTestChannels(frequency int) {
 	for {
 		time.Sleep(time.Duration(frequency) * time.Minute)
 		logger.SysLog("testing all channels")
-		_ = testAllChannels(false)
+		_ = testChannels(false, "all")
 		logger.SysLog("channel test finished")
 	}
 }

controller/channel.go

@@ -15,7 +15,7 @@ func GetAllChannels(c *gin.Context) {
 	if p < 0 {
 		p = 0
 	}
-	channels, err := model.GetAllChannels(p*config.ItemsPerPage, config.ItemsPerPage, false)
+	channels, err := model.GetAllChannels(p*config.ItemsPerPage, config.ItemsPerPage, "limited")
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,

controller/misc.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/message"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strings"
@@ -22,6 +23,7 @@ func GetStatus(c *gin.Context) {
 			"email_verification":  config.EmailVerificationEnabled,
 			"github_oauth":        config.GitHubOAuthEnabled,
 			"github_client_id":    config.GitHubClientId,
+			"lark_client_id":      config.LarkClientId,
 			"system_name":         config.SystemName,
 			"logo":                config.Logo,
 			"footer_html":         config.Footer,
@@ -110,7 +112,7 @@ func SendEmailVerification(c *gin.Context) {
 	content := fmt.Sprintf("<p>您好，你正在进行%s邮箱验证。</p>"+
 		"<p>您的验证码为: <strong>%s</strong></p>"+
 		"<p>验证码 %d 分钟内有效，如果不是本人操作，请忽略。</p>", config.SystemName, code, common.VerificationValidMinutes)
-	err := common.SendEmail(subject, email, content)
+	err := message.SendEmail(subject, email, content)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -149,7 +151,7 @@ func SendPasswordResetEmail(c *gin.Context) {
 		"<p>点击 <a href='%s'>此处</a> 进行密码重置。</p>"+
 		"<p>如果链接无法点击，请尝试点击下面的链接或将其复制到浏览器中打开：<br> %s </p>"+
 		"<p>重置链接 %d 分钟内有效，如果不是本人操作，请忽略。</p>", config.SystemName, link, link, common.VerificationValidMinutes)
-	err := common.SendEmail(subject, email, content)
+	err := message.SendEmail(subject, email, content)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,

controller/model.go

@@ -3,14 +3,15 @@ package controller
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"github.com/songquanpeng/one-api/relay/channel/ai360"
-	"github.com/songquanpeng/one-api/relay/channel/baichuan"
-	"github.com/songquanpeng/one-api/relay/channel/minimax"
-	"github.com/songquanpeng/one-api/relay/channel/mistral"
-	"github.com/songquanpeng/one-api/relay/channel/moonshot"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/helper"
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"net/http"
+	"strings"
 )
 
 // https://platform.openai.com/docs/api-reference/models/list
@@ -42,6 +43,7 @@ type OpenAIModels struct {
 
 var openAIModels []OpenAIModels
 var openAIModelsMap map[string]OpenAIModels
+var channelId2Models map[int][]string
 
 func init() {
 	var permission []OpenAIModelPermission
@@ -79,71 +81,82 @@ func init() {
 			})
 		}
 	}
-	for _, modelName := range ai360.ModelList {
-		openAIModels = append(openAIModels, OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "360",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
-		})
-	}
-	for _, modelName := range moonshot.ModelList {
-		openAIModels = append(openAIModels, OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "moonshot",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
-		})
-	}
-	for _, modelName := range baichuan.ModelList {
-		openAIModels = append(openAIModels, OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "baichuan",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
-		})
-	}
-	for _, modelName := range minimax.ModelList {
-		openAIModels = append(openAIModels, OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "minimax",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
-		})
-	}
-	for _, modelName := range mistral.ModelList {
-		openAIModels = append(openAIModels, OpenAIModels{
-			Id:         modelName,
-			Object:     "model",
-			Created:    1626777600,
-			OwnedBy:    "mistralai",
-			Permission: permission,
-			Root:       modelName,
-			Parent:     nil,
-		})
+	for _, channelType := range openai.CompatibleChannels {
+		if channelType == common.ChannelTypeAzure {
+			continue
+		}
+		channelName, channelModelList := openai.GetCompatibleChannelMeta(channelType)
+		for _, modelName := range channelModelList {
+			openAIModels = append(openAIModels, OpenAIModels{
+				Id:         modelName,
+				Object:     "model",
+				Created:    1626777600,
+				OwnedBy:    channelName,
+				Permission: permission,
+				Root:       modelName,
+				Parent:     nil,
+			})
+		}
 	}
 	openAIModelsMap = make(map[string]OpenAIModels)
 	for _, model := range openAIModels {
 		openAIModelsMap[model.Id] = model
 	}
+	channelId2Models = make(map[int][]string)
+	for i := 1; i < common.ChannelTypeDummy; i++ {
+		adaptor := helper.GetAdaptor(constant.ChannelType2APIType(i))
+		meta := &util.RelayMeta{
+			ChannelType: i,
+		}
+		adaptor.Init(meta)
+		channelId2Models[i] = adaptor.GetModelList()
+	}
+}
+
+func DashboardListModels(c *gin.Context) {
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "",
+		"data":    channelId2Models,
+	})
 }
 
 func ListModels(c *gin.Context) {
+	ctx := c.Request.Context()
+	var availableModels []string
+	if c.GetString("available_models") != "" {
+		availableModels = strings.Split(c.GetString("available_models"), ",")
+	} else {
+		userId := c.GetInt("id")
+		userGroup, _ := model.CacheGetUserGroup(userId)
+		availableModels, _ = model.CacheGetGroupModels(ctx, userGroup)
+	}
+	modelSet := make(map[string]bool)
+	for _, availableModel := range availableModels {
+		modelSet[availableModel] = true
+	}
+	availableOpenAIModels := make([]OpenAIModels, 0)
+	for _, model := range openAIModels {
+		if _, ok := modelSet[model.Id]; ok {
+			modelSet[model.Id] = false
+			availableOpenAIModels = append(availableOpenAIModels, model)
+		}
+	}
+	for modelName, ok := range modelSet {
+		if ok {
+			availableOpenAIModels = append(availableOpenAIModels, OpenAIModels{
+				Id:      modelName,
+				Object:  "model",
+				Created: 1626777600,
+				OwnedBy: "custom",
+				Root:    modelName,
+				Parent:  nil,
+			})
+		}
+	}
 	c.JSON(200, gin.H{
 		"object": "list",
-		"data":   openAIModels,
+		"data":   availableOpenAIModels,
 	})
 }
 
@@ -163,3 +176,30 @@ func RetrieveModel(c *gin.Context) {
 		})
 	}
 }
+
+func GetUserAvailableModels(c *gin.Context) {
+	ctx := c.Request.Context()
+	id := c.GetInt("id")
+	userGroup, err := model.CacheGetUserGroup(id)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	models, err := model.CacheGetGroupModels(ctx, userGroup)
+	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":    models,
+	})
+	return
+}

controller/relay.go

@@ -11,6 +11,7 @@ import (
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/middleware"
 	dbmodel "github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/monitor"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/controller"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -45,11 +46,12 @@ func Relay(c *gin.Context) {
 		requestBody, _ := common.GetRequestBody(c)
 		logger.Debugf(ctx, "request body: %s", string(requestBody))
 	}
+	channelId := c.GetInt("channel_id")
 	bizErr := relay(c, relayMode)
 	if bizErr == nil {
+		monitor.Emit(channelId, true)
 		return
 	}
-	channelId := c.GetInt("channel_id")
 	lastFailedChannelId := channelId
 	channelName := c.GetString("channel_name")
 	group := c.GetString("group")
@@ -117,7 +119,9 @@ func processChannelRelayError(ctx context.Context, channelId int, channelName st
 	logger.Errorf(ctx, "relay error (channel #%d): %s", channelId, err.Message)
 	// https://platform.openai.com/docs/guides/error-codes/api-errors
 	if util.ShouldDisableChannel(&err.Error, err.StatusCode) {
-		disableChannel(channelId, channelName, err.Message)
+		monitor.DisableChannel(channelId, channelName, err.Message)
+	} else {
+		monitor.Emit(channelId, false)
 	}
 }
 

controller/token.go

@@ -1,10 +1,12 @@
 package controller
 
 import (
+	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/network"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strconv"
@@ -16,7 +18,10 @@ func GetAllTokens(c *gin.Context) {
 	if p < 0 {
 		p = 0
 	}
-	tokens, err := model.GetAllUserTokens(userId, p*config.ItemsPerPage, config.ItemsPerPage)
+
+	order := c.Query("order")
+	tokens, err := model.GetAllUserTokens(userId, p*config.ItemsPerPage, config.ItemsPerPage, order)
+
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -101,6 +106,19 @@ func GetTokenStatus(c *gin.Context) {
 	})
 }
 
+func validateToken(c *gin.Context, token model.Token) error {
+	if len(token.Name) > 30 {
+		return fmt.Errorf("令牌名称过长")
+	}
+	if token.Subnet != nil && *token.Subnet != "" {
+		err := network.IsValidSubnets(*token.Subnet)
+		if err != nil {
+			return fmt.Errorf("无效的网段：%s", err.Error())
+		}
+	}
+	return nil
+}
+
 func AddToken(c *gin.Context) {
 	token := model.Token{}
 	err := c.ShouldBindJSON(&token)
@@ -111,13 +129,15 @@ func AddToken(c *gin.Context) {
 		})
 		return
 	}
-	if len(token.Name) > 30 {
+	err = validateToken(c, token)
+	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
-			"message": "令牌名称过长",
+			"message": fmt.Sprintf("参数错误：%s", err.Error()),
 		})
 		return
 	}
+
 	cleanToken := model.Token{
 		UserId:         c.GetInt("id"),
 		Name:           token.Name,
@@ -127,6 +147,8 @@ func AddToken(c *gin.Context) {
 		ExpiredTime:    token.ExpiredTime,
 		RemainQuota:    token.RemainQuota,
 		UnlimitedQuota: token.UnlimitedQuota,
+		Models:         token.Models,
+		Subnet:         token.Subnet,
 	}
 	err = cleanToken.Insert()
 	if err != nil {
@@ -139,6 +161,7 @@ func AddToken(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
+		"data":    cleanToken,
 	})
 	return
 }
@@ -173,10 +196,11 @@ func UpdateToken(c *gin.Context) {
 		})
 		return
 	}
-	if len(token.Name) > 30 {
+	err = validateToken(c, token)
+	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
-			"message": "令牌名称过长",
+			"message": fmt.Sprintf("参数错误：%s", err.Error()),
 		})
 		return
 	}
@@ -212,6 +236,8 @@ func UpdateToken(c *gin.Context) {
 		cleanToken.ExpiredTime = token.ExpiredTime
 		cleanToken.RemainQuota = token.RemainQuota
 		cleanToken.UnlimitedQuota = token.UnlimitedQuota
+		cleanToken.Models = token.Models
+		cleanToken.Subnet = token.Subnet
 	}
 	err = cleanToken.Update()
 	if err != nil {

controller/user.go

@@ -58,11 +58,11 @@ func Login(c *gin.Context) {
 		})
 		return
 	}
-	setupLogin(&user, c)
+	SetupLogin(&user, c)
 }
 
 // setup session & cookies and then return user info
-func setupLogin(user *model.User, c *gin.Context) {
+func SetupLogin(user *model.User, c *gin.Context) {
 	session := sessions.Default(c)
 	session.Set("id", user.Id)
 	session.Set("username", user.Username)
@@ -184,7 +184,10 @@ func GetAllUsers(c *gin.Context) {
 	if p < 0 {
 		p = 0
 	}
-	users, err := model.GetAllUsers(p*config.ItemsPerPage, config.ItemsPerPage)
+
+	order := c.DefaultQuery("order", "")
+	users, err := model.GetAllUsers(p*config.ItemsPerPage, config.ItemsPerPage, order)
+
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -192,12 +195,12 @@ func GetAllUsers(c *gin.Context) {
 		})
 		return
 	}
+
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
 		"data":    users,
 	})
-	return
 }
 
 func SearchUsers(c *gin.Context) {
@@ -767,3 +770,38 @@ func TopUp(c *gin.Context) {
 	})
 	return
 }
+
+type adminTopUpRequest struct {
+	UserId int    `json:"user_id"`
+	Quota  int    `json:"quota"`
+	Remark string `json:"remark"`
+}
+
+func AdminTopUp(c *gin.Context) {
+	req := adminTopUpRequest{}
+	err := c.ShouldBindJSON(&req)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	err = model.IncreaseUserQuota(req.UserId, int64(req.Quota))
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+	if req.Remark == "" {
+		req.Remark = fmt.Sprintf("通过 API 充值 %s", common.LogQuota(int64(req.Quota)))
+	}
+	model.RecordTopupLog(req.UserId, req.Remark, req.Quota)
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "",
+	})
+	return
+}

docker-compose.yml

@@ -2,7 +2,7 @@ version: '3.4'
 
 services:
   one-api:
-    image: justsong/one-api:latest
+    image: "${REGISTRY:-docker.io}/justsong/one-api:latest"
     container_name: one-api
     restart: always
     command: --log-dir /app/logs
@@ -29,12 +29,12 @@ services:
       retries: 3
 
   redis:
-    image: redis:latest
+    image: "${REGISTRY:-docker.io}/redis:latest"
     container_name: redis
     restart: always
 
   db:
-    image: mysql:8.2.0
+    image: "${REGISTRY:-docker.io}/mysql:8.2.0"
     restart: always
     container_name: mysql
     volumes:

docs/API.md

@@ -0,0 +1,53 @@
+# 使用 API 操控 & 扩展 One API
+> 欢迎提交 PR 在此放上你的拓展项目。
+
+例如，虽然 One API 本身没有直接支持支付，但是你可以通过系统扩展的 API 来实现支付功能。
+
+又或者你想自定义渠道管理策略，也可以通过 API 来实现渠道的禁用与启用。
+
+## 鉴权
+One API 支持两种鉴权方式：Cookie 和 Token，对于 Token，参照下图获取：
+
+![image](https://github.com/songquanpeng/songquanpeng.github.io/assets/39998050/c15281a7-83ed-47cb-a1f6-913cb6bf4a7c)
+
+之后，将 Token 作为请求头的 Authorization 字段的值即可，例如下面使用 Token 调用测试渠道的 API：
+![image](https://github.com/songquanpeng/songquanpeng.github.io/assets/39998050/1273b7ae-cb60-4c0d-93a6-b1cbc039c4f8)
+
+## 请求格式与响应格式
+One API 使用 JSON 格式进行请求和响应。
+
+对于响应体，一般格式如下：
+```json
+{
+  "message": "请求信息",
+  "success": true,
+  "data": {}
+}
+```
+
+## API 列表
+> 当前 API 列表不全，请自行通过浏览器抓取前端请求
+
+如果现有的 API 没有办法满足你的需求，欢迎提交 issue 讨论。
+
+### 获取当前登录用户信息
+**GET** `/api/user/self`
+
+### 为给定用户充值额度
+**POST** `/api/topup`
+```json
+{
+  "user_id": 1,
+  "quota": 100000,
+  "remark": "充值 100000 额度"
+}
+```
+
+## 其他
+### 充值链接上的附加参数
+One API 会在用户点击充值按钮的时候，将用户的信息和充值信息附加在链接上，例如：
+`https://example.com?username=root&user_id=1&transaction_id=4b3eed80-55d5-443f-bd44-fb18c648c837`
+
+你可以通过解析链接上的参数来获取用户信息和充值信息，然后调用 API 来为用户充值。
+
+注意，不是所有主题都支持该功能，欢迎 PR 补齐。

go.mod

@@ -15,6 +15,7 @@ require (
 	github.com/google/uuid v1.3.0
 	github.com/gorilla/websocket v1.5.0
 	github.com/pkoukk/tiktoken-go v0.1.5
+	github.com/smartystreets/goconvey v1.8.1
 	github.com/stretchr/testify v1.8.3
 	golang.org/x/crypto v0.17.0
 	golang.org/x/image v0.14.0
@@ -37,15 +38,18 @@ require (
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-sql-driver/mysql v1.6.0 // indirect
 	github.com/goccy/go-json v0.10.2 // indirect
+	github.com/gopherjs/gopherjs v1.17.2 // indirect
 	github.com/gorilla/context v1.1.1 // indirect
 	github.com/gorilla/securecookie v1.1.1 // indirect
 	github.com/gorilla/sessions v1.2.1 // indirect
 	github.com/jackc/pgpassfile v1.0.0 // indirect
 	github.com/jackc/pgservicefile v0.0.0-20221227161230-091c0ba34f0a // indirect
-	github.com/jackc/pgx/v5 v5.3.1 // indirect
+	github.com/jackc/pgx/v5 v5.5.4 // indirect
+	github.com/jackc/puddle/v2 v2.2.1 // indirect
 	github.com/jinzhu/inflection v1.0.0 // indirect
 	github.com/jinzhu/now v1.1.5 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
+	github.com/jtolds/gls v4.20.0+incompatible // indirect
 	github.com/klauspost/cpuid/v2 v2.2.4 // indirect
 	github.com/leodido/go-urn v1.2.4 // indirect
 	github.com/mattn/go-isatty v0.0.19 // indirect
@@ -54,12 +58,14 @@ require (
 	github.com/modern-go/reflect2 v1.0.2 // indirect
 	github.com/pelletier/go-toml/v2 v2.0.8 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
+	github.com/smarty/assertions v1.15.0 // indirect
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.11 // indirect
 	golang.org/x/arch v0.3.0 // indirect
 	golang.org/x/net v0.17.0 // indirect
+	golang.org/x/sync v0.1.0 // indirect
 	golang.org/x/sys v0.15.0 // indirect
 	golang.org/x/text v0.14.0 // indirect
-	google.golang.org/protobuf v1.30.0 // indirect
+	google.golang.org/protobuf v1.33.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )

go.sum

@@ -56,11 +56,13 @@ github.com/golang-jwt/jwt v3.2.2+incompatible h1:IfV12K8xAKAnZqdXVzCZ+TOjboZ2keL
 github.com/golang-jwt/jwt v3.2.2+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
 github.com/golang/protobuf v1.3.3/go.mod h1:vzj43D7+SQXF/4pzW/hwtAqwc6iTitCiVSaWz5lYuqw=
 github.com/golang/protobuf v1.5.0/go.mod h1:FsONVRAS9T7sI+LIUmWTfcYkHO4aIWwzhcaSAoJOfIk=
-github.com/google/go-cmp v0.5.5 h1:Khx7svrCpmxxtHBq5j2mp/xVjsi8hQMfNLvJFAlrGgU=
 github.com/google/go-cmp v0.5.5/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
+github.com/google/go-cmp v0.5.6 h1:BKbKCqvP6I+rmFHt06ZmyQtvB8xAkWdhFyr0ZUNZcxQ=
 github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
 github.com/google/uuid v1.3.0 h1:t6JiXgmwXMjEs8VusXIJk2BXHsn+wx8BZdTaoZ5fu7I=
 github.com/google/uuid v1.3.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
+github.com/gopherjs/gopherjs v1.17.2 h1:fQnZVsXk8uxXIStYb0N4bGk7jeyTalG/wsZjQ25dO0g=
+github.com/gopherjs/gopherjs v1.17.2/go.mod h1:pRRIvn/QzFLrKfvEz3qUuEhtE/zLCWfreZ6J5gM2i+k=
 github.com/gorilla/context v1.1.1 h1:AWwleXJkX/nhcU9bZSnZoi3h/qGYqQAGhq6zZe/aQW8=
 github.com/gorilla/context v1.1.1/go.mod h1:kBGZzfjB9CEq2AlWe17Uuf7NDRt0dE0s8S51q0aT7Yg=
 github.com/gorilla/securecookie v1.1.1 h1:miw7JPhV+b/lAHSXz4qd/nN9jRiAFV5FwjeKyCS8BvQ=
@@ -73,8 +75,10 @@ github.com/jackc/pgpassfile v1.0.0 h1:/6Hmqy13Ss2zCq62VdNG8tM1wchn8zjSGOBJ6icpsI
 github.com/jackc/pgpassfile v1.0.0/go.mod h1:CEx0iS5ambNFdcRtxPj5JhEz+xB6uRky5eyVu/W2HEg=
 github.com/jackc/pgservicefile v0.0.0-20221227161230-091c0ba34f0a h1:bbPeKD0xmW/Y25WS6cokEszi5g+S0QxI/d45PkRi7Nk=
 github.com/jackc/pgservicefile v0.0.0-20221227161230-091c0ba34f0a/go.mod h1:5TJZWKEWniPve33vlWYSoGYefn3gLQRzjfDlhSJ9ZKM=
-github.com/jackc/pgx/v5 v5.3.1 h1:Fcr8QJ1ZeLi5zsPZqQeUZhNhxfkkKBOgJuYkJHoBOtU=
-github.com/jackc/pgx/v5 v5.3.1/go.mod h1:t3JDKnCBlYIc0ewLF0Q7B8MXmoIaBOZj/ic7iHozM/8=
+github.com/jackc/pgx/v5 v5.5.4 h1:Xp2aQS8uXButQdnCMWNmvx6UysWQQC+u1EoizjguY+8=
+github.com/jackc/pgx/v5 v5.5.4/go.mod h1:ez9gk+OAat140fv9ErkZDYFWmXLfV+++K0uAOiwgm1A=
+github.com/jackc/puddle/v2 v2.2.1 h1:RhxXJtFG022u4ibrCSMSiu5aOq1i77R3OHKNJj77OAk=
+github.com/jackc/puddle/v2 v2.2.1/go.mod h1:vriiEXHvEE654aYKXXjOvZM39qJ0q+azkZFrfEOc3H4=
 github.com/jinzhu/inflection v1.0.0 h1:K317FqzuhWc8YvSVlFMCCUb36O/S9MCKRDI7QkRKD/E=
 github.com/jinzhu/inflection v1.0.0/go.mod h1:h+uFLlag+Qp1Va5pdKtLDYj+kHp5pxUVkryuEj+Srlc=
 github.com/jinzhu/now v1.1.4/go.mod h1:d3SSVoowX0Lcu0IBviAWJpolVfI5UJVZZ7cO71lE/z8=
@@ -83,6 +87,8 @@ github.com/jinzhu/now v1.1.5/go.mod h1:d3SSVoowX0Lcu0IBviAWJpolVfI5UJVZZ7cO71lE/
 github.com/json-iterator/go v1.1.9/go.mod h1:KdQUCv79m/52Kvf8AW2vK1V8akMuk1QjK/uOdHXbAo4=
 github.com/json-iterator/go v1.1.12 h1:PV8peI4a0ysnczrg+LtxykD8LfKY9ML6u2jnxaEnrnM=
 github.com/json-iterator/go v1.1.12/go.mod h1:e30LSqwooZae/UwlEbR2852Gd8hjQvJoHmT4TnhNGBo=
+github.com/jtolds/gls v4.20.0+incompatible h1:xdiiI2gbIgH/gLH7ADydsJ1uDOEzR8yvV7C0MuV77Wo=
+github.com/jtolds/gls v4.20.0+incompatible/go.mod h1:QJZ7F/aHp+rZTRtaJ1ow/lLfFfVYBRgL+9YlvaHOwJU=
 github.com/klauspost/cpuid/v2 v2.0.9/go.mod h1:FInQzS24/EEf25PyTYn52gqo7WaD8xa0213Md/qVLRg=
 github.com/klauspost/cpuid/v2 v2.2.4 h1:acbojRNwl3o09bUq+yDCtZFc1aiwaAAxtcn8YkZXnvk=
 github.com/klauspost/cpuid/v2 v2.2.4/go.mod h1:RVVoqg1df56z8g3pUjL/3lE5UfnlrJX8tyFgg4nqhuY=
@@ -125,6 +131,10 @@ github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZN
 github.com/rogpeppe/go-internal v1.6.1/go.mod h1:xXDCJY+GAPziupqXw64V24skbSoqbTEfhy4qGm1nDQc=
 github.com/rogpeppe/go-internal v1.8.0 h1:FCbCCtXNOY3UtUuHUYaghJg4y7Fd14rXifAYUAtL9R8=
 github.com/rogpeppe/go-internal v1.8.0/go.mod h1:WmiCO8CzOY8rg0OYDC4/i/2WRWAB6poM+XZ2dLUbcbE=
+github.com/smarty/assertions v1.15.0 h1:cR//PqUBUiQRakZWqBiFFQ9wb8emQGDb0HeGdqGByCY=
+github.com/smarty/assertions v1.15.0/go.mod h1:yABtdzeQs6l1brC900WlRNwj6ZR55d7B+E8C6HtKdec=
+github.com/smartystreets/goconvey v1.8.1 h1:qGjIddxOk4grTu9JPOU31tVfq3cNdBlNa5sSznIX1xY=
+github.com/smartystreets/goconvey v1.8.1/go.mod h1:+/u4qLyY6x1jReYOp7GOM2FSt8aP9CzCZL03bI28W60=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/objx v0.4.0/go.mod h1:YvHI0jy2hoMjB+UWwv71VJQ9isScKT/TqJzVSSt89Yw=
 github.com/stretchr/objx v0.5.0/go.mod h1:Yh+to48EsGEfYuaHDzXPcE3xhTkx73EhmCGUpEOglKo=
@@ -157,6 +167,8 @@ golang.org/x/image v0.14.0/go.mod h1:HUYqC05R2ZcZ3ejNQsIHQDQiwWM4JBqmm6MKANTp4LE
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=
 golang.org/x/net v0.17.0 h1:pVaXccu2ozPjCXewfr1S7xza/zcXTity9cCdXQYSjIM=
 golang.org/x/net v0.17.0/go.mod h1:NxSsAGuq816PNPmqtQdLE42eU2Fs7NoRIZrHJAlaCOE=
+golang.org/x/sync v0.1.0 h1:wsuoTGHzEhffawBOhz5CYhcrV4IdKZbEyZjBMuTp12o=
+golang.org/x/sync v0.1.0/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sys v0.0.0-20200116001909-b77594299b42/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20210615035016-665e8c7367d1/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
@@ -173,12 +185,12 @@ golang.org/x/text v0.3.6/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.14.0 h1:ScX5w1eTa3QqT8oi6+ziP7dTV1S2+ALU0bI+0zXKWiQ=
 golang.org/x/text v0.14.0/go.mod h1:18ZOQIKpY8NJVqYksKHtTdi31H5itFRjB5/qKTNYzSU=
 golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
-golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543 h1:E7g+9GITq07hpfrRu66IVDexMakfv52eLZ2CXBWiKr4=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 h1:go1bK/D/BFZV2I8cIQd1NKEZ+0owSTG1fDTci4IqFcE=
 google.golang.org/protobuf v1.26.0-rc.1/go.mod h1:jlhhOSvTdKEhbULTjvd4ARK9grFBp09yW+WbY/TyQbw=
 google.golang.org/protobuf v1.28.0/go.mod h1:HV8QOd/L58Z+nl8r43ehVNZIU/HEI6OcFqwMG9pJV4I=
-google.golang.org/protobuf v1.30.0 h1:kPPoIgf3TsEvrm0PFe15JQ+570QVxYzEvvHqChK+cng=
-google.golang.org/protobuf v1.30.0/go.mod h1:HV8QOd/L58Z+nl8r43ehVNZIU/HEI6OcFqwMG9pJV4I=
+google.golang.org/protobuf v1.33.0 h1:uNO2rsAINq/JlFpSdYEKIZ0uKD/R9cpdv0T+yoGwGmI=
+google.golang.org/protobuf v1.33.0/go.mod h1:c6P6GXX6sHbq/GpV6MGZEdwhWPcYBgnhAHhKbcUYpos=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20180628173108-788fd7840127/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=

i18n/en.json

@@ -8,12 +8,12 @@
   "确认删除": "Confirm Delete",
   "确认绑定": "Confirm Binding",
   "您正在删除自己的帐户，将清空所有数据且不可恢复": "You are deleting your account, all data will be cleared and unrecoverable.",
-  "\"通道「%s」（#%d）已被禁用\"": "\"Channel %s (#%d) has been disabled\"",
-  "通道「%s」（#%d）已被禁用，原因：%s": "Channel %s (#%d) has been disabled, reason: %s",
+  "\"渠道「%s」（#%d）已被禁用\"": "\"Channel %s (#%d) has been disabled\"",
+  "渠道「%s」（#%d）已被禁用，原因：%s": "Channel %s (#%d) has been disabled, reason: %s",
   "测试已在运行中": "Test is already running",
   "响应时间 %.2fs 超过阈值 %.2fs": "Response time %.2fs exceeds threshold %.2fs",
-  "通道测试完成": "Channel test completed",
-  "通道测试完成，如果没有收到禁用通知，说明所有通道都正常": "Channel test completed, if you have not received the disable notification, it means that all channels are normal",
+  "渠道测试完成": "Channel test completed",
+  "渠道测试完成，如果没有收到禁用通知，说明所有渠道都正常": "Channel test completed, if you have not received the disable notification, it means that all channels are normal",
   "无法连接至 GitHub 服务器，请稍后重试！": "Unable to connect to GitHub server, please try again later!",
   "返回值非法，用户字段为空，请稍后重试！": "The return value is illegal, the user field is empty, please try again later!",
   "管理员未开启通过 GitHub 登录以及注册": "The administrator did not turn on login and registration via GitHub",
@@ -119,11 +119,11 @@
   " 个月 ": " M ",
   " 年 ": " y ",
   "未测试": "Not tested",
-  "通道 ${name} 测试成功，耗时 ${time.toFixed(2)} 秒。": "Channel ${name} test succeeded, time consumed ${time.toFixed(2)} s.",
-  "已成功开始测试所有通道，请刷新页面查看结果。": "All channels have been successfully tested, please refresh the page to view the results.",
-  "已成功开始测试所有已启用通道，请刷新页面查看结果。": "All enabled channels have been successfully tested, please refresh the page to view the results.",
-  "通道 ${name} 余额更新成功！": "Channel ${name} balance updated successfully!",
-  "已更新完毕所有已启用通道余额！": "The balance of all enabled channels has been updated!",
+  "渠道 ${name} 测试成功，耗时 ${time.toFixed(2)} 秒。": "Channel ${name} test succeeded, time consumed ${time.toFixed(2)} s.",
+  "已成功开始测试所有渠道，请刷新页面查看结果。": "All channels have been successfully tested, please refresh the page to view the results.",
+  "已成功开始测试所有已启用渠道，请刷新页面查看结果。": "All enabled channels have been successfully tested, please refresh the page to view the results.",
+  "渠道 ${name} 余额更新成功！": "Channel ${name} balance updated successfully!",
+  "已更新完毕所有已启用渠道余额！": "The balance of all enabled channels has been updated!",
   "搜索渠道的 ID，名称和密钥 ...": "Search for channel ID, name and key ...",
   "名称": "Name",
   "分组": "Group",
@@ -141,9 +141,9 @@
   "启用": "Enable",
   "编辑": "Edit",
   "添加新的渠道": "Add a new channel",
-  "测试所有通道": "Test all channels",
-  "测试所有已启用通道": "Test all enabled channels",
-  "更新所有已启用通道余额": "Update the balance of all enabled channels",
+  "测试所有渠道": "Test all channels",
+  "测试所有已启用渠道": "Test all enabled channels",
+  "更新所有已启用渠道余额": "Update the balance of all enabled channels",
   "刷新": "Refresh",
   "处理中...": "Processing...",
   "绑定成功！": "Binding succeeded!",
@@ -207,11 +207,11 @@
   "监控设置": "Monitoring Settings",
   "最长响应时间": "Longest Response Time",
   "单位秒": "Unit in seconds",
-  "当运行通道全部测试时": "When all operating channels are tested",
-  "超过此时间将自动禁用通道": "Channels will be automatically disabled if this time is exceeded",
+  "当运行渠道全部测试时": "When all operating channels are tested",
+  "超过此时间将自动禁用渠道": "Channels will be automatically disabled if this time is exceeded",
   "额度提醒阈值": "Quota reminder threshold",
   "低于此额度时将发送邮件提醒用户": "Email will be sent to remind users when the quota is below this",
-  "失败时自动禁用通道": "Automatically disable the channel when it fails",
+  "失败时自动禁用渠道": "Automatically disable the channel when it fails",
   "保存监控设置": "Save Monitoring Settings",
   "额度设置": "Quota Settings",
   "新用户初始额度": "Initial quota for new users",
@@ -405,7 +405,7 @@
   "镜像": "Mirror",
   "请输入镜像站地址，格式为：https://domain.com，可不填，不填则使用渠道默认值": "Please enter the mirror site address, the format is: https://domain.com, it can be left blank, if left blank, the default value of the channel will be used",
   "模型": "Model",
-  "请选择该通道所支持的模型": "Please select the model supported by the channel",
+  "请选择该渠道所支持的模型": "Please select the model supported by the channel",
   "填入基础模型": "Fill in the basic model",
   "填入所有模型": "Fill in all models",
   "清除所有模型": "Clear all models",
@@ -515,7 +515,7 @@
   "请输入自定义渠道的 Base URL": "Please enter the Base URL of the custom channel",
   "Homepage URL 填": "Fill in the Homepage URL",
   "Authorization callback URL 填": "Fill in the Authorization callback URL",
-  "请为通道命名": "Please name the channel",
+  "请为渠道命名": "Please name the channel",
   "此项可选，用于修改请求体中的模型名称，为一个 JSON 字符串，键为请求中模型名称，值为要替换的模型名称，例如：": "This is optional, used to modify the model name in the request body, it's a JSON string, the key is the model name in the request, and the value is the model name to be replaced, for example:",
   "模型重定向": "Model redirection",
   "请输入渠道对应的鉴权密钥": "Please enter the authentication key corresponding to the channel",

main.go

@@ -30,11 +30,25 @@ func main() {
 	if config.DebugEnabled {
 		logger.SysLog("running in debug mode")
 	}
+	var err error
 	// Initialize SQL Database
-	err := model.InitDB()
+	model.DB, err = model.InitDB("SQL_DSN")
 	if err != nil {
 		logger.FatalLog("failed to initialize database: " + err.Error())
 	}
+	if os.Getenv("LOG_SQL_DSN") != "" {
+		logger.SysLog("using secondary database for table logs")
+		model.LOG_DB, err = model.InitDB("LOG_SQL_DSN")
+		if err != nil {
+			logger.FatalLog("failed to initialize secondary database: " + err.Error())
+		}
+	} else {
+		model.LOG_DB = model.DB
+	}
+	err = model.CreateRootAccountIfNeed()
+	if err != nil {
+		logger.FatalLog("database init error: " + err.Error())
+	}
 	defer func() {
 		err := model.CloseDB()
 		if err != nil {
@@ -64,13 +78,6 @@ func main() {
 		go model.SyncOptions(config.SyncFrequency)
 		go model.SyncChannelCache(config.SyncFrequency)
 	}
-	if os.Getenv("CHANNEL_UPDATE_FREQUENCY") != "" {
-		frequency, err := strconv.Atoi(os.Getenv("CHANNEL_UPDATE_FREQUENCY"))
-		if err != nil {
-			logger.FatalLog("failed to parse CHANNEL_UPDATE_FREQUENCY: " + err.Error())
-		}
-		go controller.AutomaticallyUpdateChannels(frequency)
-	}
 	if os.Getenv("CHANNEL_TEST_FREQUENCY") != "" {
 		frequency, err := strconv.Atoi(os.Getenv("CHANNEL_TEST_FREQUENCY"))
 		if err != nil {
@@ -83,6 +90,9 @@ func main() {
 		logger.SysLog("batch update enabled with interval " + strconv.Itoa(config.BatchUpdateInterval) + "s")
 		model.InitBatchUpdater()
 	}
+	if config.EnableMetric {
+		logger.SysLog("metric enabled, will disable channel if too much request failed")
+	}
 	openai.InitTokenEncoders()
 
 	// Initialize HTTP server

middleware/auth.go

@@ -1,9 +1,12 @@
 package middleware
 
 import (
+	"fmt"
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/blacklist"
+	"github.com/songquanpeng/one-api/common/network"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strings"
@@ -42,11 +45,14 @@ func authHelper(c *gin.Context, minRole int) {
 			return
 		}
 	}
-	if status.(int) == common.UserStatusDisabled {
+	if status.(int) == common.UserStatusDisabled || blacklist.IsUserBanned(id.(int)) {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
 			"message": "用户已被封禁",
 		})
+		session := sessions.Default(c)
+		session.Clear()
+		_ = session.Save()
 		c.Abort()
 		return
 	}
@@ -84,6 +90,7 @@ func RootAuth() func(c *gin.Context) {
 
 func TokenAuth() func(c *gin.Context) {
 	return func(c *gin.Context) {
+		ctx := c.Request.Context()
 		key := c.Request.Header.Get("Authorization")
 		key = strings.TrimPrefix(key, "Bearer ")
 		key = strings.TrimPrefix(key, "sk-")
@@ -94,15 +101,34 @@ func TokenAuth() func(c *gin.Context) {
 			abortWithMessage(c, http.StatusUnauthorized, err.Error())
 			return
 		}
+		if token.Subnet != nil && *token.Subnet != "" {
+			if !network.IsIpInSubnets(ctx, c.ClientIP(), *token.Subnet) {
+				abortWithMessage(c, http.StatusForbidden, fmt.Sprintf("该令牌只能在指定网段使用：%s，当前 ip：%s", *token.Subnet, c.ClientIP()))
+				return
+			}
+		}
 		userEnabled, err := model.CacheIsUserEnabled(token.UserId)
 		if err != nil {
 			abortWithMessage(c, http.StatusInternalServerError, err.Error())
 			return
 		}
-		if !userEnabled {
+		if !userEnabled || blacklist.IsUserBanned(token.UserId) {
 			abortWithMessage(c, http.StatusForbidden, "用户已被封禁")
 			return
 		}
+		requestModel, err := getRequestModel(c)
+		if err != nil && !strings.HasPrefix(c.Request.URL.Path, "/v1/models") {
+			abortWithMessage(c, http.StatusBadRequest, err.Error())
+			return
+		}
+		c.Set("request_model", requestModel)
+		if token.Models != nil && *token.Models != "" {
+			c.Set("available_models", *token.Models)
+			if requestModel != "" && !isModelInList(requestModel, *token.Models) {
+				abortWithMessage(c, http.StatusForbidden, fmt.Sprintf("该令牌无权使用模型：%s", requestModel))
+				return
+			}
+		}
 		c.Set("id", token.UserId)
 		c.Set("token_id", token.Id)
 		c.Set("token_name", token.Name)

middleware/distributor.go

@@ -2,14 +2,12 @@ package middleware
 
 import (
 	"fmt"
+	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/model"
 	"net/http"
 	"strconv"
-	"strings"
-
-	"github.com/gin-gonic/gin"
 )
 
 type ModelRequest struct {
@@ -40,37 +38,11 @@ func Distribute() func(c *gin.Context) {
 				return
 			}
 		} else {
-			// Select a channel for the user
-			var modelRequest ModelRequest
-			err := common.UnmarshalBodyReusable(c, &modelRequest)
+			requestModel := c.GetString("request_model")
+			var err error
+			channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, requestModel, false)
 			if err != nil {
-				abortWithMessage(c, http.StatusBadRequest, "无效的请求")
-				return
-			}
-			if strings.HasPrefix(c.Request.URL.Path, "/v1/moderations") {
-				if modelRequest.Model == "" {
-					modelRequest.Model = "text-moderation-stable"
-				}
-			}
-			if strings.HasSuffix(c.Request.URL.Path, "embeddings") {
-				if modelRequest.Model == "" {
-					modelRequest.Model = c.Param("model")
-				}
-			}
-			if strings.HasPrefix(c.Request.URL.Path, "/v1/images/generations") {
-				if modelRequest.Model == "" {
-					modelRequest.Model = "dall-e-2"
-				}
-			}
-			if strings.HasPrefix(c.Request.URL.Path, "/v1/audio/transcriptions") || strings.HasPrefix(c.Request.URL.Path, "/v1/audio/translations") {
-				if modelRequest.Model == "" {
-					modelRequest.Model = "whisper-1"
-				}
-			}
-			requestModel = modelRequest.Model
-			channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, modelRequest.Model, false)
-			if err != nil {
-				message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", userGroup, modelRequest.Model)
+				message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", userGroup, requestModel)
 				if channel != nil {
 					logger.SysError(fmt.Sprintf("渠道不存在：%d", channel.Id))
 					message = "数据库一致性已被破坏，请联系管理员"

middleware/recover.go

@@ -3,6 +3,7 @@ package middleware
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/logger"
 	"net/http"
 	"runtime/debug"
@@ -12,11 +13,15 @@ func RelayPanicRecover() gin.HandlerFunc {
 	return func(c *gin.Context) {
 		defer func() {
 			if err := recover(); err != nil {
-				logger.SysError(fmt.Sprintf("panic detected: %v", err))
-				logger.SysError(fmt.Sprintf("stacktrace from panic: %s", string(debug.Stack())))
+				ctx := c.Request.Context()
+				logger.Errorf(ctx, fmt.Sprintf("panic detected: %v", err))
+				logger.Errorf(ctx, fmt.Sprintf("stacktrace from panic: %s", string(debug.Stack())))
+				logger.Errorf(ctx, fmt.Sprintf("request: %s %s", c.Request.Method, c.Request.URL.Path))
+				body, _ := common.GetRequestBody(c)
+				logger.Errorf(ctx, fmt.Sprintf("request body: %s", string(body)))
 				c.JSON(http.StatusInternalServerError, gin.H{
 					"error": gin.H{
-						"message": fmt.Sprintf("Panic detected, error: %v. Please submit a issue here: https://github.com/songquanpeng/one-api", err),
+						"message": fmt.Sprintf("Panic detected, error: %v. Please submit an issue with the related log here: https://github.com/songquanpeng/one-api", err),
 						"type":    "one_api_panic",
 					},
 				})

middleware/request-id.go

@@ -9,7 +9,7 @@ import (
 
 func RequestId() func(c *gin.Context) {
 	return func(c *gin.Context) {
-		id := helper.GetTimeString() + helper.GetRandomNumberString(8)
+		id := helper.GenRequestID()
 		c.Set(logger.RequestIdKey, id)
 		ctx := context.WithValue(c.Request.Context(), logger.RequestIdKey, id)
 		c.Request = c.Request.WithContext(ctx)

middleware/utils.go

@@ -1,9 +1,12 @@
 package middleware
 
 import (
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
+	"strings"
 )
 
 func abortWithMessage(c *gin.Context, statusCode int, message string) {
@@ -16,3 +19,42 @@ func abortWithMessage(c *gin.Context, statusCode int, message string) {
 	c.Abort()
 	logger.Error(c.Request.Context(), message)
 }
+
+func getRequestModel(c *gin.Context) (string, error) {
+	var modelRequest ModelRequest
+	err := common.UnmarshalBodyReusable(c, &modelRequest)
+	if err != nil {
+		return "", fmt.Errorf("common.UnmarshalBodyReusable failed: %w", err)
+	}
+	if strings.HasPrefix(c.Request.URL.Path, "/v1/moderations") {
+		if modelRequest.Model == "" {
+			modelRequest.Model = "text-moderation-stable"
+		}
+	}
+	if strings.HasSuffix(c.Request.URL.Path, "embeddings") {
+		if modelRequest.Model == "" {
+			modelRequest.Model = c.Param("model")
+		}
+	}
+	if strings.HasPrefix(c.Request.URL.Path, "/v1/images/generations") {
+		if modelRequest.Model == "" {
+			modelRequest.Model = "dall-e-2"
+		}
+	}
+	if strings.HasPrefix(c.Request.URL.Path, "/v1/audio/transcriptions") || strings.HasPrefix(c.Request.URL.Path, "/v1/audio/translations") {
+		if modelRequest.Model == "" {
+			modelRequest.Model = "whisper-1"
+		}
+	}
+	return modelRequest.Model, nil
+}
+
+func isModelInList(modelName string, models string) bool {
+	modelList := strings.Split(models, ",")
+	for _, model := range modelList {
+		if modelName == model {
+			return true
+		}
+	}
+	return false
+}

model/ability.go

@@ -1,7 +1,10 @@
 package model
 
 import (
+	"context"
 	"github.com/songquanpeng/one-api/common"
+	"gorm.io/gorm"
+	"sort"
 	"strings"
 )
 
@@ -13,7 +16,7 @@ type Ability struct {
 	Priority  *int64 `json:"priority" gorm:"bigint;default:0;index"`
 }
 
-func GetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
+func GetRandomSatisfiedChannel(group string, model string, ignoreFirstPriority bool) (*Channel, error) {
 	ability := Ability{}
 	groupCol := "`group`"
 	trueVal := "1"
@@ -23,8 +26,13 @@ func GetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
 	}
 
 	var err error = nil
-	maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model)
-	channelQuery := DB.Where(groupCol+" = ? and model = ? and enabled = "+trueVal+" and priority = (?)", group, model, maxPrioritySubQuery)
+	var channelQuery *gorm.DB
+	if ignoreFirstPriority {
+		channelQuery = DB.Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model)
+	} else {
+		maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model)
+		channelQuery = DB.Where(groupCol+" = ? and model = ? and enabled = "+trueVal+" and priority = (?)", group, model, maxPrioritySubQuery)
+	}
 	if common.UsingSQLite || common.UsingPostgreSQL {
 		err = channelQuery.Order("RANDOM()").First(&ability).Error
 	} else {
@@ -82,3 +90,19 @@ func (channel *Channel) UpdateAbilities() error {
 func UpdateAbilityStatus(channelId int, status bool) error {
 	return DB.Model(&Ability{}).Where("channel_id = ?", channelId).Select("enabled").Update("enabled", status).Error
 }
+
+func GetGroupModels(ctx context.Context, group string) ([]string, error) {
+	groupCol := "`group`"
+	trueVal := "1"
+	if common.UsingPostgreSQL {
+		groupCol = `"group"`
+		trueVal = "true"
+	}
+	var models []string
+	err := DB.Model(&Ability{}).Distinct("model").Where(groupCol+" = ? and enabled = "+trueVal, group).Pluck("model", &models).Error
+	if err != nil {
+		return nil, err
+	}
+	sort.Strings(models)
+	return models, err
+}

model/cache.go

@@ -1,6 +1,7 @@
 package model
 
 import (
+	"context"
 	"encoding/json"
 	"errors"
 	"fmt"
@@ -20,6 +21,7 @@ var (
 	UserId2GroupCacheSeconds  = config.SyncFrequency
 	UserId2QuotaCacheSeconds  = config.SyncFrequency
 	UserId2StatusCacheSeconds = config.SyncFrequency
+	GroupModelsCacheSeconds   = config.SyncFrequency
 )
 
 func CacheGetTokenByKey(key string) (*Token, error) {
@@ -70,31 +72,42 @@ func CacheGetUserGroup(id int) (group string, err error) {
 	return group, err
 }
 
-func CacheGetUserQuota(id int) (quota int, err error) {
+func fetchAndUpdateUserQuota(ctx context.Context, id int) (quota int64, err error) {
+	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 {
+		logger.Error(ctx, "Redis set user quota error: "+err.Error())
+	}
+	return
+}
+
+func CacheGetUserQuota(ctx context.Context, id int) (quota int64, 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 {
-			logger.SysError("Redis set user quota error: " + err.Error())
-		}
-		return quota, err
+		return fetchAndUpdateUserQuota(ctx, id)
 	}
-	quota, err = strconv.Atoi(quotaString)
-	return quota, err
+	quota, err = strconv.ParseInt(quotaString, 10, 64)
+	if err != nil {
+		return 0, nil
+	}
+	if quota <= config.PreConsumedQuota { // when user's quota is less than pre-consumed quota, we need to fetch from db
+		logger.Infof(ctx, "user %d's cached quota is too low: %d, refreshing from db", quota, id)
+		return fetchAndUpdateUserQuota(ctx, id)
+	}
+	return quota, nil
 }
 
-func CacheUpdateUserQuota(id int) error {
+func CacheUpdateUserQuota(ctx context.Context, id int) error {
 	if !common.RedisEnabled {
 		return nil
 	}
-	quota, err := CacheGetUserQuota(id)
+	quota, err := CacheGetUserQuota(ctx, id)
 	if err != nil {
 		return err
 	}
@@ -102,7 +115,7 @@ func CacheUpdateUserQuota(id int) error {
 	return err
 }
 
-func CacheDecreaseUserQuota(id int, quota int) error {
+func CacheDecreaseUserQuota(id int, quota int64) error {
 	if !common.RedisEnabled {
 		return nil
 	}
@@ -134,6 +147,25 @@ func CacheIsUserEnabled(userId int) (bool, error) {
 	return userEnabled, err
 }
 
+func CacheGetGroupModels(ctx context.Context, group string) ([]string, error) {
+	if !common.RedisEnabled {
+		return GetGroupModels(ctx, group)
+	}
+	modelsStr, err := common.RedisGet(fmt.Sprintf("group_models:%s", group))
+	if err == nil {
+		return strings.Split(modelsStr, ","), nil
+	}
+	models, err := GetGroupModels(ctx, group)
+	if err != nil {
+		return nil, err
+	}
+	err = common.RedisSet(fmt.Sprintf("group_models:%s", group), strings.Join(models, ","), time.Duration(GroupModelsCacheSeconds)*time.Second)
+	if err != nil {
+		logger.SysError("Redis set group models error: " + err.Error())
+	}
+	return models, nil
+}
+
 var group2model2channels map[string]map[string][]*Channel
 var channelSyncLock sync.RWMutex
 
@@ -193,7 +225,7 @@ func SyncChannelCache(frequency int) {
 
 func CacheGetRandomSatisfiedChannel(group string, model string, ignoreFirstPriority bool) (*Channel, error) {
 	if !config.MemoryCacheEnabled {
-		return GetRandomSatisfiedChannel(group, model)
+		return GetRandomSatisfiedChannel(group, model, ignoreFirstPriority)
 	}
 	channelSyncLock.RLock()
 	defer channelSyncLock.RUnlock()

model/channel.go

@@ -13,7 +13,7 @@ import (
 type Channel struct {
 	Id                 int     `json:"id"`
 	Type               int     `json:"type" gorm:"default:0"`
-	Key                string  `json:"key" gorm:"not null;index"`
+	Key                string  `json:"key" gorm:"type:text"`
 	Status             int     `json:"status" gorm:"default:1"`
 	Name               string  `json:"name" gorm:"index"`
 	Weight             *uint   `json:"weight" gorm:"default:0"`
@@ -32,23 +32,22 @@ type Channel struct {
 	Config             string  `json:"config"`
 }
 
-func GetAllChannels(startIdx int, num int, selectAll bool) ([]*Channel, error) {
+func GetAllChannels(startIdx int, num int, scope string) ([]*Channel, error) {
 	var channels []*Channel
 	var err error
-	if selectAll {
+	switch scope {
+	case "all":
 		err = DB.Order("id desc").Find(&channels).Error
-	} else {
+	case "disabled":
+		err = DB.Order("id desc").Where("status = ? or status = ?", common.ChannelStatusAutoDisabled, common.ChannelStatusManuallyDisabled).Find(&channels).Error
+	default:
 		err = DB.Order("id desc").Limit(num).Offset(startIdx).Omit("key").Find(&channels).Error
 	}
 	return channels, err
 }
 
 func SearchChannels(keyword string) (channels []*Channel, err error) {
-	keyCol := "`key`"
-	if common.UsingPostgreSQL {
-		keyCol = `"key"`
-	}
-	err = DB.Omit("key").Where("id = ? or name LIKE ? or "+keyCol+" = ?", helper.String2Int(keyword), keyword+"%", keyword).Find(&channels).Error
+	err = DB.Omit("key").Where("id = ? or name LIKE ?", helper.String2Int(keyword), keyword+"%").Find(&channels).Error
 	return channels, err
 }
 
@@ -179,7 +178,7 @@ func UpdateChannelStatusById(id int, status int) {
 	}
 }
 
-func UpdateChannelUsedQuota(id int, quota int) {
+func UpdateChannelUsedQuota(id int, quota int64) {
 	if config.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeChannelUsedQuota, id, quota)
 		return
@@ -187,7 +186,7 @@ func UpdateChannelUsedQuota(id int, quota int) {
 	updateChannelUsedQuota(id, quota)
 }
 
-func updateChannelUsedQuota(id int, quota int) {
+func updateChannelUsedQuota(id int, quota int64) {
 	err := DB.Model(&Channel{}).Where("id = ?", id).Update("used_quota", gorm.Expr("used_quota + ?", quota)).Error
 	if err != nil {
 		logger.SysError("failed to update channel used quota: " + err.Error())

model/log.go

@@ -45,13 +45,28 @@ func RecordLog(userId int, logType int, content string) {
 		Type:      logType,
 		Content:   content,
 	}
-	err := DB.Create(log).Error
+	err := LOG_DB.Create(log).Error
 	if err != nil {
 		logger.SysError("failed to record log: " + err.Error())
 	}
 }
 
-func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptTokens int, completionTokens int, modelName string, tokenName string, quota int, content string) {
+func RecordTopupLog(userId int, content string, quota int) {
+	log := &Log{
+		UserId:    userId,
+		Username:  GetUsernameById(userId),
+		CreatedAt: helper.GetTimestamp(),
+		Type:      LogTypeTopup,
+		Content:   content,
+		Quota:     quota,
+	}
+	err := LOG_DB.Create(log).Error
+	if err != nil {
+		logger.SysError("failed to record log: " + err.Error())
+	}
+}
+
+func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptTokens int, completionTokens int, modelName string, tokenName string, quota int64, content string) {
 	logger.Info(ctx, fmt.Sprintf("record consume log: userId=%d, channelId=%d, promptTokens=%d, completionTokens=%d, modelName=%s, tokenName=%s, quota=%d, content=%s", userId, channelId, promptTokens, completionTokens, modelName, tokenName, quota, content))
 	if !config.LogConsumeEnabled {
 		return
@@ -66,10 +81,10 @@ func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptToke
 		CompletionTokens: completionTokens,
 		TokenName:        tokenName,
 		ModelName:        modelName,
-		Quota:            quota,
+		Quota:            int(quota),
 		ChannelId:        channelId,
 	}
-	err := DB.Create(log).Error
+	err := LOG_DB.Create(log).Error
 	if err != nil {
 		logger.Error(ctx, "failed to record log: "+err.Error())
 	}
@@ -78,9 +93,9 @@ func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptToke
 func GetAllLogs(logType int, startTimestamp int64, endTimestamp int64, modelName string, username string, tokenName string, startIdx int, num int, channel int) (logs []*Log, err error) {
 	var tx *gorm.DB
 	if logType == LogTypeUnknown {
-		tx = DB
+		tx = LOG_DB
 	} else {
-		tx = DB.Where("type = ?", logType)
+		tx = LOG_DB.Where("type = ?", logType)
 	}
 	if modelName != "" {
 		tx = tx.Where("model_name = ?", modelName)
@@ -107,9 +122,9 @@ 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) (logs []*Log, err error) {
 	var tx *gorm.DB
 	if logType == LogTypeUnknown {
-		tx = DB.Where("user_id = ?", userId)
+		tx = LOG_DB.Where("user_id = ?", userId)
 	} else {
-		tx = DB.Where("user_id = ? and type = ?", userId, logType)
+		tx = LOG_DB.Where("user_id = ? and type = ?", userId, logType)
 	}
 	if modelName != "" {
 		tx = tx.Where("model_name = ?", modelName)
@@ -128,17 +143,17 @@ func GetUserLogs(userId int, logType int, startTimestamp int64, endTimestamp int
 }
 
 func SearchAllLogs(keyword string) (logs []*Log, err error) {
-	err = DB.Where("type = ? or content LIKE ?", keyword, keyword+"%").Order("id desc").Limit(config.MaxRecentItems).Find(&logs).Error
+	err = LOG_DB.Where("type = ? or content LIKE ?", keyword, keyword+"%").Order("id desc").Limit(config.MaxRecentItems).Find(&logs).Error
 	return logs, err
 }
 
 func SearchUserLogs(userId int, keyword string) (logs []*Log, err error) {
-	err = DB.Where("user_id = ? and type = ?", userId, keyword).Order("id desc").Limit(config.MaxRecentItems).Omit("id").Find(&logs).Error
+	err = LOG_DB.Where("user_id = ? and type = ?", userId, keyword).Order("id desc").Limit(config.MaxRecentItems).Omit("id").Find(&logs).Error
 	return logs, err
 }
 
-func SumUsedQuota(logType int, startTimestamp int64, endTimestamp int64, modelName string, username string, tokenName string, channel int) (quota int) {
-	tx := DB.Table("logs").Select("ifnull(sum(quota),0)")
+func SumUsedQuota(logType int, startTimestamp int64, endTimestamp int64, modelName string, username string, tokenName string, channel int) (quota int64) {
+	tx := LOG_DB.Table("logs").Select("ifnull(sum(quota),0)")
 	if username != "" {
 		tx = tx.Where("username = ?", username)
 	}
@@ -162,7 +177,7 @@ func SumUsedQuota(logType int, startTimestamp int64, endTimestamp int64, modelNa
 }
 
 func SumUsedToken(logType int, startTimestamp int64, endTimestamp int64, modelName string, username string, tokenName string) (token int) {
-	tx := DB.Table("logs").Select("ifnull(sum(prompt_tokens),0) + ifnull(sum(completion_tokens),0)")
+	tx := LOG_DB.Table("logs").Select("ifnull(sum(prompt_tokens),0) + ifnull(sum(completion_tokens),0)")
 	if username != "" {
 		tx = tx.Where("username = ?", username)
 	}
@@ -183,7 +198,7 @@ func SumUsedToken(logType int, startTimestamp int64, endTimestamp int64, modelNa
 }
 
 func DeleteOldLog(targetTimestamp int64) (int64, error) {
-	result := DB.Where("created_at < ?", targetTimestamp).Delete(&Log{})
+	result := LOG_DB.Where("created_at < ?", targetTimestamp).Delete(&Log{})
 	return result.RowsAffected, result.Error
 }
 
@@ -207,7 +222,7 @@ func SearchLogsByDayAndModel(userId, start, end int) (LogStatistics []*LogStatis
 		groupSelect = "strftime('%Y-%m-%d', datetime(created_at, 'unixepoch')) as day"
 	}
 
-	err = DB.Raw(`
+	err = LOG_DB.Raw(`
 		SELECT `+groupSelect+`,
 		model_name, count(1) as request_count,
 		sum(quota) as quota,

model/main.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/env"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
 	"gorm.io/driver/mysql"
@@ -16,12 +17,13 @@ import (
 )
 
 var DB *gorm.DB
+var LOG_DB *gorm.DB
 
-func createRootAccountIfNeed() error {
+func CreateRootAccountIfNeed() error {
 	var user User
 	//if user.Status != util.UserStatusEnabled {
 	if err := DB.First(&user).Error; err != nil {
-		logger.SysLog("no user exists, create a root user for you: username is root, password is 123456")
+		logger.SysLog("no user exists, creating a root user for you: username is root, password is 123456")
 		hashedPassword, err := common.Password2Hash("123456")
 		if err != nil {
 			return err
@@ -33,16 +35,32 @@ func createRootAccountIfNeed() error {
 			Status:      common.UserStatusEnabled,
 			DisplayName: "Root User",
 			AccessToken: helper.GetUUID(),
-			Quota:       100000000,
+			Quota:       500000000000000,
 		}
 		DB.Create(&rootUser)
+		if config.InitialRootToken != "" {
+			logger.SysLog("creating initial root token as requested")
+			token := Token{
+				Id:             1,
+				UserId:         rootUser.Id,
+				Key:            config.InitialRootToken,
+				Status:         common.TokenStatusEnabled,
+				Name:           "Initial Root Token",
+				CreatedTime:    helper.GetTimestamp(),
+				AccessedTime:   helper.GetTimestamp(),
+				ExpiredTime:    -1,
+				RemainQuota:    500000000000000,
+				UnlimitedQuota: true,
+			}
+			DB.Create(&token)
+		}
 	}
 	return nil
 }
 
-func chooseDB() (*gorm.DB, error) {
-	if os.Getenv("SQL_DSN") != "" {
-		dsn := os.Getenv("SQL_DSN")
+func chooseDB(envName string) (*gorm.DB, error) {
+	if os.Getenv(envName) != "" {
+		dsn := os.Getenv(envName)
 		if strings.HasPrefix(dsn, "postgres://") {
 			// Use PostgreSQL
 			logger.SysLog("using PostgreSQL as database")
@@ -56,6 +74,7 @@ func chooseDB() (*gorm.DB, error) {
 		}
 		// Use MySQL
 		logger.SysLog("using MySQL as database")
+		common.UsingMySQL = true
 		return gorm.Open(mysql.Open(dsn), &gorm.Config{
 			PrepareStmt: true, // precompile SQL
 		})
@@ -69,67 +88,78 @@ func chooseDB() (*gorm.DB, error) {
 	})
 }
 
-func InitDB() (err error) {
-	db, err := chooseDB()
+func InitDB(envName string) (db *gorm.DB, err error) {
+	db, err = chooseDB(envName)
 	if err == nil {
-		if config.DebugEnabled {
+		if config.DebugSQLEnabled {
 			db = db.Debug()
 		}
-		DB = db
-		sqlDB, err := DB.DB()
+		sqlDB, err := db.DB()
 		if err != nil {
-			return err
+			return nil, err
 		}
-		sqlDB.SetMaxIdleConns(helper.GetOrDefaultEnvInt("SQL_MAX_IDLE_CONNS", 100))
-		sqlDB.SetMaxOpenConns(helper.GetOrDefaultEnvInt("SQL_MAX_OPEN_CONNS", 1000))
-		sqlDB.SetConnMaxLifetime(time.Second * time.Duration(helper.GetOrDefaultEnvInt("SQL_MAX_LIFETIME", 60)))
+		sqlDB.SetMaxIdleConns(env.Int("SQL_MAX_IDLE_CONNS", 100))
+		sqlDB.SetMaxOpenConns(env.Int("SQL_MAX_OPEN_CONNS", 1000))
+		sqlDB.SetConnMaxLifetime(time.Second * time.Duration(env.Int("SQL_MAX_LIFETIME", 60)))
 
 		if !config.IsMasterNode {
-			return nil
+			return db, err
+		}
+		if common.UsingMySQL {
+			_, _ = sqlDB.Exec("DROP INDEX idx_channels_key ON channels;") // TODO: delete this line when most users have upgraded
 		}
 		logger.SysLog("database migration started")
 		err = db.AutoMigrate(&Channel{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&Token{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&User{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&Option{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&Redemption{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&Ability{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		err = db.AutoMigrate(&Log{})
 		if err != nil {
-			return err
+			return nil, err
 		}
 		logger.SysLog("database migrated")
-		err = createRootAccountIfNeed()
-		return err
+		return db, err
 	} else {
 		logger.FatalLog(err)
 	}
-	return err
+	return db, err
 }
 
-func CloseDB() error {
-	sqlDB, err := DB.DB()
+func closeDB(db *gorm.DB) error {
+	sqlDB, err := db.DB()
 	if err != nil {
 		return err
 	}
 	err = sqlDB.Close()
 	return err
 }
+
+func CloseDB() error {
+	if LOG_DB != DB {
+		err := closeDB(LOG_DB)
+		if err != nil {
+			return err
+		}
+	}
+	return closeDB(DB)
+}

model/option.go

@@ -57,13 +57,15 @@ func InitOptionMap() {
 	config.OptionMap["WeChatServerAddress"] = ""
 	config.OptionMap["WeChatServerToken"] = ""
 	config.OptionMap["WeChatAccountQRCodeImageURL"] = ""
+	config.OptionMap["MessagePusherAddress"] = ""
+	config.OptionMap["MessagePusherToken"] = ""
 	config.OptionMap["TurnstileSiteKey"] = ""
 	config.OptionMap["TurnstileSecretKey"] = ""
-	config.OptionMap["QuotaForNewUser"] = strconv.Itoa(config.QuotaForNewUser)
-	config.OptionMap["QuotaForInviter"] = strconv.Itoa(config.QuotaForInviter)
-	config.OptionMap["QuotaForInvitee"] = strconv.Itoa(config.QuotaForInvitee)
-	config.OptionMap["QuotaRemindThreshold"] = strconv.Itoa(config.QuotaRemindThreshold)
-	config.OptionMap["PreConsumedQuota"] = strconv.Itoa(config.PreConsumedQuota)
+	config.OptionMap["QuotaForNewUser"] = strconv.FormatInt(config.QuotaForNewUser, 10)
+	config.OptionMap["QuotaForInviter"] = strconv.FormatInt(config.QuotaForInviter, 10)
+	config.OptionMap["QuotaForInvitee"] = strconv.FormatInt(config.QuotaForInvitee, 10)
+	config.OptionMap["QuotaRemindThreshold"] = strconv.FormatInt(config.QuotaRemindThreshold, 10)
+	config.OptionMap["PreConsumedQuota"] = strconv.FormatInt(config.PreConsumedQuota, 10)
 	config.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
 	config.OptionMap["GroupRatio"] = common.GroupRatio2JSONString()
 	config.OptionMap["CompletionRatio"] = common.CompletionRatio2JSONString()
@@ -79,6 +81,9 @@ func InitOptionMap() {
 func loadOptionsFromDatabase() {
 	options, _ := AllOption()
 	for _, option := range options {
+		if option.Key == "ModelRatio" {
+			option.Value = common.AddNewMissingRatio(option.Value)
+		}
 		err := updateOptionMap(option.Key, option.Value)
 		if err != nil {
 			logger.SysError("failed to update option map: " + err.Error())
@@ -167,6 +172,10 @@ func updateOptionMap(key string, value string) (err error) {
 		config.GitHubClientId = value
 	case "GitHubClientSecret":
 		config.GitHubClientSecret = value
+	case "LarkClientId":
+		config.LarkClientId = value
+	case "LarkClientSecret":
+		config.LarkClientSecret = value
 	case "Footer":
 		config.Footer = value
 	case "SystemName":
@@ -179,20 +188,24 @@ func updateOptionMap(key string, value string) (err error) {
 		config.WeChatServerToken = value
 	case "WeChatAccountQRCodeImageURL":
 		config.WeChatAccountQRCodeImageURL = value
+	case "MessagePusherAddress":
+		config.MessagePusherAddress = value
+	case "MessagePusherToken":
+		config.MessagePusherToken = value
 	case "TurnstileSiteKey":
 		config.TurnstileSiteKey = value
 	case "TurnstileSecretKey":
 		config.TurnstileSecretKey = value
 	case "QuotaForNewUser":
-		config.QuotaForNewUser, _ = strconv.Atoi(value)
+		config.QuotaForNewUser, _ = strconv.ParseInt(value, 10, 64)
 	case "QuotaForInviter":
-		config.QuotaForInviter, _ = strconv.Atoi(value)
+		config.QuotaForInviter, _ = strconv.ParseInt(value, 10, 64)
 	case "QuotaForInvitee":
-		config.QuotaForInvitee, _ = strconv.Atoi(value)
+		config.QuotaForInvitee, _ = strconv.ParseInt(value, 10, 64)
 	case "QuotaRemindThreshold":
-		config.QuotaRemindThreshold, _ = strconv.Atoi(value)
+		config.QuotaRemindThreshold, _ = strconv.ParseInt(value, 10, 64)
 	case "PreConsumedQuota":
-		config.PreConsumedQuota, _ = strconv.Atoi(value)
+		config.PreConsumedQuota, _ = strconv.ParseInt(value, 10, 64)
 	case "RetryTimes":
 		config.RetryTimes, _ = strconv.Atoi(value)
 	case "ModelRatio":

model/redemption.go

@@ -14,7 +14,7 @@ type Redemption struct {
 	Key          string `json:"key" gorm:"type:char(32);uniqueIndex"`
 	Status       int    `json:"status" gorm:"default:1"`
 	Name         string `json:"name" gorm:"index"`
-	Quota        int    `json:"quota" gorm:"default:100"`
+	Quota        int64  `json:"quota" gorm:"bigint;default:100"`
 	CreatedTime  int64  `json:"created_time" gorm:"bigint"`
 	RedeemedTime int64  `json:"redeemed_time" gorm:"bigint"`
 	Count        int    `json:"count" gorm:"-:all"` // only for api request
@@ -42,7 +42,7 @@ func GetRedemptionById(id int) (*Redemption, error) {
 	return &redemption, err
 }
 
-func Redeem(key string, userId int) (quota int, err error) {
+func Redeem(key string, userId int) (quota int64, err error) {
 	if key == "" {
 		return 0, errors.New("未提供兑换码")
 	}

model/token.go

@@ -7,27 +7,41 @@ import (
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/message"
 	"gorm.io/gorm"
 )
 
 type Token struct {
-	Id             int    `json:"id"`
-	UserId         int    `json:"user_id"`
-	Key            string `json:"key" gorm:"type:char(48);uniqueIndex"`
-	Status         int    `json:"status" gorm:"default:1"`
-	Name           string `json:"name" gorm:"index" `
-	CreatedTime    int64  `json:"created_time" gorm:"bigint"`
-	AccessedTime   int64  `json:"accessed_time" gorm:"bigint"`
-	ExpiredTime    int64  `json:"expired_time" gorm:"bigint;default:-1"` // -1 means never expired
-	RemainQuota    int    `json:"remain_quota" gorm:"default:0"`
-	UnlimitedQuota bool   `json:"unlimited_quota" gorm:"default:false"`
-	UsedQuota      int    `json:"used_quota" gorm:"default:0"` // used quota
+	Id             int     `json:"id"`
+	UserId         int     `json:"user_id"`
+	Key            string  `json:"key" gorm:"type:char(48);uniqueIndex"`
+	Status         int     `json:"status" gorm:"default:1"`
+	Name           string  `json:"name" gorm:"index" `
+	CreatedTime    int64   `json:"created_time" gorm:"bigint"`
+	AccessedTime   int64   `json:"accessed_time" gorm:"bigint"`
+	ExpiredTime    int64   `json:"expired_time" gorm:"bigint;default:-1"` // -1 means never expired
+	RemainQuota    int64   `json:"remain_quota" gorm:"bigint;default:0"`
+	UnlimitedQuota bool    `json:"unlimited_quota" gorm:"default:false"`
+	UsedQuota      int64   `json:"used_quota" gorm:"bigint;default:0"` // used quota
+	Models         *string `json:"models" gorm:"default:''"`           // allowed models
+	Subnet         *string `json:"subnet" gorm:"default:''"`           // allowed subnet
 }
 
-func GetAllUserTokens(userId int, startIdx int, num int) ([]*Token, error) {
+func GetAllUserTokens(userId int, startIdx int, num int, order string) ([]*Token, error) {
 	var tokens []*Token
 	var err error
-	err = DB.Where("user_id = ?", userId).Order("id desc").Limit(num).Offset(startIdx).Find(&tokens).Error
+	query := DB.Where("user_id = ?", userId)
+
+	switch order {
+	case "remain_quota":
+		query = query.Order("unlimited_quota desc, remain_quota desc")
+	case "used_quota":
+		query = query.Order("used_quota desc")
+	default:
+		query = query.Order("id desc")
+	}
+
+	err = query.Limit(num).Offset(startIdx).Find(&tokens).Error
 	return tokens, err
 }
 
@@ -49,7 +63,7 @@ func ValidateUserToken(key string) (token *Token, err error) {
 		return nil, errors.New("令牌验证失败")
 	}
 	if token.Status == common.TokenStatusExhausted {
-		return nil, errors.New("该令牌额度已用尽")
+		return nil, fmt.Errorf("令牌 %s（#%d）额度已用尽", token.Name, token.Id)
 	} else if token.Status == common.TokenStatusExpired {
 		return nil, errors.New("该令牌已过期")
 	}
@@ -109,7 +123,7 @@ 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
-	err = DB.Model(token).Select("name", "status", "expired_time", "remain_quota", "unlimited_quota").Updates(token).Error
+	err = DB.Model(token).Select("name", "status", "expired_time", "remain_quota", "unlimited_quota", "models", "subnet").Updates(token).Error
 	return err
 }
 
@@ -137,7 +151,7 @@ func DeleteTokenById(id int, userId int) (err error) {
 	return token.Delete()
 }
 
-func IncreaseTokenQuota(id int, quota int) (err error) {
+func IncreaseTokenQuota(id int, quota int64) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
@@ -148,7 +162,7 @@ func IncreaseTokenQuota(id int, quota int) (err error) {
 	return increaseTokenQuota(id, quota)
 }
 
-func increaseTokenQuota(id int, quota int) (err error) {
+func increaseTokenQuota(id int, quota int64) (err error) {
 	err = DB.Model(&Token{}).Where("id = ?", id).Updates(
 		map[string]interface{}{
 			"remain_quota":  gorm.Expr("remain_quota + ?", quota),
@@ -159,7 +173,7 @@ func increaseTokenQuota(id int, quota int) (err error) {
 	return err
 }
 
-func DecreaseTokenQuota(id int, quota int) (err error) {
+func DecreaseTokenQuota(id int, quota int64) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
@@ -170,7 +184,7 @@ func DecreaseTokenQuota(id int, quota int) (err error) {
 	return decreaseTokenQuota(id, quota)
 }
 
-func decreaseTokenQuota(id int, quota int) (err error) {
+func decreaseTokenQuota(id int, quota int64) (err error) {
 	err = DB.Model(&Token{}).Where("id = ?", id).Updates(
 		map[string]interface{}{
 			"remain_quota":  gorm.Expr("remain_quota - ?", quota),
@@ -181,7 +195,7 @@ func decreaseTokenQuota(id int, quota int) (err error) {
 	return err
 }
 
-func PreConsumeTokenQuota(tokenId int, quota int) (err error) {
+func PreConsumeTokenQuota(tokenId int, quota int64) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
@@ -213,7 +227,7 @@ func PreConsumeTokenQuota(tokenId int, quota int) (err error) {
 			}
 			if email != "" {
 				topUpLink := fmt.Sprintf("%s/topup", config.ServerAddress)
-				err = common.SendEmail(prompt, email,
+				err = message.SendEmail(prompt, email,
 					fmt.Sprintf("%s，当前剩余额度为 %d，为了不影响您的使用，请及时充值。<br/>充值链接：<a href='%s'>%s</a>", prompt, userQuota, topUpLink, topUpLink))
 				if err != nil {
 					logger.SysError("failed to send email" + err.Error())
@@ -231,7 +245,7 @@ func PreConsumeTokenQuota(tokenId int, quota int) (err error) {
 	return err
 }
 
-func PostConsumeTokenQuota(tokenId int, quota int) (err error) {
+func PostConsumeTokenQuota(tokenId int, quota int64) (err error) {
 	token, err := GetTokenById(tokenId)
 	if quota > 0 {
 		err = DecreaseUserQuota(token.UserId, quota)

model/user.go

@@ -4,6 +4,7 @@ import (
 	"errors"
 	"fmt"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/blacklist"
 	"github.com/songquanpeng/one-api/common/config"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
@@ -23,11 +24,12 @@ type User struct {
 	Email            string `json:"email" gorm:"index" validate:"max=50"`
 	GitHubId         string `json:"github_id" gorm:"column:github_id;index"`
 	WeChatId         string `json:"wechat_id" gorm:"column:wechat_id;index"`
+	LarkId           string `json:"lark_id" gorm:"column:lark_id;index"`
 	VerificationCode string `json:"verification_code" gorm:"-:all"`                                    // this field is only for Email verification, don't save it to database!
 	AccessToken      string `json:"access_token" gorm:"type:char(32);column:access_token;uniqueIndex"` // this token is for system management
-	Quota            int    `json:"quota" gorm:"type:int;default:0"`
-	UsedQuota        int    `json:"used_quota" gorm:"type:int;default:0;column:used_quota"` // used quota
-	RequestCount     int    `json:"request_count" gorm:"type:int;default:0;"`               // request number
+	Quota            int64  `json:"quota" gorm:"bigint;default:0"`
+	UsedQuota        int64  `json:"used_quota" gorm:"bigint;default:0;column:used_quota"` // used quota
+	RequestCount     int    `json:"request_count" gorm:"type:int;default:0;"`             // request number
 	Group            string `json:"group" gorm:"type:varchar(32);default:'default'"`
 	AffCode          string `json:"aff_code" gorm:"type:varchar(32);column:aff_code;uniqueIndex"`
 	InviterId        int    `json:"inviter_id" gorm:"type:int;column:inviter_id;index"`
@@ -39,8 +41,21 @@ func GetMaxUserId() int {
 	return user.Id
 }
 
-func GetAllUsers(startIdx int, num int) (users []*User, err error) {
-	err = DB.Order("id desc").Limit(num).Offset(startIdx).Omit("password").Find(&users).Error
+func GetAllUsers(startIdx int, num int, order string) (users []*User, err error) {
+	query := DB.Limit(num).Offset(startIdx).Omit("password").Where("status != ?", common.UserStatusDeleted)
+
+	switch order {
+	case "quota":
+		query = query.Order("quota desc")
+	case "used_quota":
+		query = query.Order("used_quota desc")
+	case "request_count":
+		query = query.Order("request_count desc")
+	default:
+		query = query.Order("id desc")
+	}
+
+	err = query.Find(&users).Error
 	return users, err
 }
 
@@ -123,6 +138,11 @@ func (user *User) Update(updatePassword bool) error {
 			return err
 		}
 	}
+	if user.Status == common.UserStatusDisabled {
+		blacklist.BanUser(user.Id)
+	} else if user.Status == common.UserStatusEnabled {
+		blacklist.UnbanUser(user.Id)
+	}
 	err = DB.Model(user).Updates(user).Error
 	return err
 }
@@ -131,7 +151,10 @@ func (user *User) Delete() error {
 	if user.Id == 0 {
 		return errors.New("id 为空！")
 	}
-	err := DB.Delete(user).Error
+	blacklist.BanUser(user.Id)
+	user.Username = fmt.Sprintf("deleted_%s", helper.GetUUID())
+	user.Status = common.UserStatusDeleted
+	err := DB.Model(user).Updates(user).Error
 	return err
 }
 
@@ -184,6 +207,14 @@ func (user *User) FillUserByGitHubId() error {
 	return nil
 }
 
+func (user *User) FillUserByLarkId() error {
+	if user.LarkId == "" {
+		return errors.New("lark id 为空！")
+	}
+	DB.Where(User{LarkId: user.LarkId}).First(user)
+	return nil
+}
+
 func (user *User) FillUserByWeChatId() error {
 	if user.WeChatId == "" {
 		return errors.New("WeChat id 为空！")
@@ -212,6 +243,10 @@ func IsGitHubIdAlreadyTaken(githubId string) bool {
 	return DB.Where("github_id = ?", githubId).Find(&User{}).RowsAffected == 1
 }
 
+func IsLarkIdAlreadyTaken(githubId string) bool {
+	return DB.Where("lark_id = ?", githubId).Find(&User{}).RowsAffected == 1
+}
+
 func IsUsernameAlreadyTaken(username string) bool {
 	return DB.Where("username = ?", username).Find(&User{}).RowsAffected == 1
 }
@@ -265,12 +300,12 @@ func ValidateAccessToken(token string) (user *User) {
 	return nil
 }
 
-func GetUserQuota(id int) (quota int, err error) {
+func GetUserQuota(id int) (quota int64, err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Select("quota").Find(&quota).Error
 	return quota, err
 }
 
-func GetUserUsedQuota(id int) (quota int, err error) {
+func GetUserUsedQuota(id int) (quota int64, err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Select("used_quota").Find(&quota).Error
 	return quota, err
 }
@@ -290,7 +325,7 @@ func GetUserGroup(id int) (group string, err error) {
 	return group, err
 }
 
-func IncreaseUserQuota(id int, quota int) (err error) {
+func IncreaseUserQuota(id int, quota int64) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
@@ -301,12 +336,12 @@ func IncreaseUserQuota(id int, quota int) (err error) {
 	return increaseUserQuota(id, quota)
 }
 
-func increaseUserQuota(id int, quota int) (err error) {
+func increaseUserQuota(id int, quota int64) (err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Update("quota", gorm.Expr("quota + ?", quota)).Error
 	return err
 }
 
-func DecreaseUserQuota(id int, quota int) (err error) {
+func DecreaseUserQuota(id int, quota int64) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
 	}
@@ -317,7 +352,7 @@ func DecreaseUserQuota(id int, quota int) (err error) {
 	return decreaseUserQuota(id, quota)
 }
 
-func decreaseUserQuota(id int, quota int) (err error) {
+func decreaseUserQuota(id int, quota int64) (err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Update("quota", gorm.Expr("quota - ?", quota)).Error
 	return err
 }
@@ -327,7 +362,7 @@ func GetRootUserEmail() (email string) {
 	return email
 }
 
-func UpdateUserUsedQuotaAndRequestCount(id int, quota int) {
+func UpdateUserUsedQuotaAndRequestCount(id int, quota int64) {
 	if config.BatchUpdateEnabled {
 		addNewRecord(BatchUpdateTypeUsedQuota, id, quota)
 		addNewRecord(BatchUpdateTypeRequestCount, id, 1)
@@ -336,7 +371,7 @@ func UpdateUserUsedQuotaAndRequestCount(id int, quota int) {
 	updateUserUsedQuotaAndRequestCount(id, quota, 1)
 }
 
-func updateUserUsedQuotaAndRequestCount(id int, quota int, count int) {
+func updateUserUsedQuotaAndRequestCount(id int, quota int64, count int) {
 	err := DB.Model(&User{}).Where("id = ?", id).Updates(
 		map[string]interface{}{
 			"used_quota":    gorm.Expr("used_quota + ?", quota),
@@ -348,7 +383,7 @@ func updateUserUsedQuotaAndRequestCount(id int, quota int, count int) {
 	}
 }
 
-func updateUserUsedQuota(id int, quota int) {
+func updateUserUsedQuota(id int, quota int64) {
 	err := DB.Model(&User{}).Where("id = ?", id).Updates(
 		map[string]interface{}{
 			"used_quota": gorm.Expr("used_quota + ?", quota),

model/utils.go

@@ -16,12 +16,12 @@ const (
 	BatchUpdateTypeCount // if you add a new type, you need to add a new map and a new lock
 )
 
-var batchUpdateStores []map[int]int
+var batchUpdateStores []map[int]int64
 var batchUpdateLocks []sync.Mutex
 
 func init() {
 	for i := 0; i < BatchUpdateTypeCount; i++ {
-		batchUpdateStores = append(batchUpdateStores, make(map[int]int))
+		batchUpdateStores = append(batchUpdateStores, make(map[int]int64))
 		batchUpdateLocks = append(batchUpdateLocks, sync.Mutex{})
 	}
 }
@@ -35,7 +35,7 @@ func InitBatchUpdater() {
 	}()
 }
 
-func addNewRecord(type_ int, id int, value int) {
+func addNewRecord(type_ int, id int, value int64) {
 	batchUpdateLocks[type_].Lock()
 	defer batchUpdateLocks[type_].Unlock()
 	if _, ok := batchUpdateStores[type_][id]; !ok {
@@ -50,7 +50,7 @@ func batchUpdate() {
 	for i := 0; i < BatchUpdateTypeCount; i++ {
 		batchUpdateLocks[i].Lock()
 		store := batchUpdateStores[i]
-		batchUpdateStores[i] = make(map[int]int)
+		batchUpdateStores[i] = make(map[int]int64)
 		batchUpdateLocks[i].Unlock()
 		// TODO: maybe we can combine updates with same key?
 		for key, value := range store {
@@ -68,7 +68,7 @@ func batchUpdate() {
 			case BatchUpdateTypeUsedQuota:
 				updateUserUsedQuota(key, value)
 			case BatchUpdateTypeRequestCount:
-				updateUserRequestCount(key, value)
+				updateUserRequestCount(key, int(value))
 			case BatchUpdateTypeChannelUsedQuota:
 				updateChannelUsedQuota(key, value)
 			}

monitor/channel.go

@@ -0,0 +1,55 @@
+package monitor
+
+import (
+	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/message"
+	"github.com/songquanpeng/one-api/model"
+)
+
+func notifyRootUser(subject string, content string) {
+	if config.MessagePusherAddress != "" {
+		err := message.SendMessage(subject, content, content)
+		if err != nil {
+			logger.SysError(fmt.Sprintf("failed to send message: %s", err.Error()))
+		} else {
+			return
+		}
+	}
+	if config.RootUserEmail == "" {
+		config.RootUserEmail = model.GetRootUserEmail()
+	}
+	err := message.SendEmail(subject, config.RootUserEmail, content)
+	if err != nil {
+		logger.SysError(fmt.Sprintf("failed to send email: %s", err.Error()))
+	}
+}
+
+// DisableChannel disable & notify
+func DisableChannel(channelId int, channelName string, reason string) {
+	model.UpdateChannelStatusById(channelId, common.ChannelStatusAutoDisabled)
+	logger.SysLog(fmt.Sprintf("channel #%d has been disabled: %s", channelId, reason))
+	subject := fmt.Sprintf("渠道「%s」（#%d）已被禁用", channelName, channelId)
+	content := fmt.Sprintf("渠道「%s」（#%d）已被禁用，原因：%s", channelName, channelId, reason)
+	notifyRootUser(subject, content)
+}
+
+func MetricDisableChannel(channelId int, successRate float64) {
+	model.UpdateChannelStatusById(channelId, common.ChannelStatusAutoDisabled)
+	logger.SysLog(fmt.Sprintf("channel #%d has been disabled due to low success rate: %.2f", channelId, successRate*100))
+	subject := fmt.Sprintf("渠道 #%d 已被禁用", channelId)
+	content := fmt.Sprintf("该渠道（#%d）在最近 %d 次调用中成功率为 %.2f%%，低于阈值 %.2f%%，因此被系统自动禁用。",
+		channelId, config.MetricQueueSize, successRate*100, config.MetricSuccessRateThreshold*100)
+	notifyRootUser(subject, content)
+}
+
+// EnableChannel enable & notify
+func EnableChannel(channelId int, channelName string) {
+	model.UpdateChannelStatusById(channelId, common.ChannelStatusEnabled)
+	logger.SysLog(fmt.Sprintf("channel #%d has been enabled", channelId))
+	subject := fmt.Sprintf("渠道「%s」（#%d）已被启用", channelName, channelId)
+	content := fmt.Sprintf("渠道「%s」（#%d）已被启用", channelName, channelId)
+	notifyRootUser(subject, content)
+}

monitor/metric.go

@@ -0,0 +1,79 @@
+package monitor
+
+import (
+	"github.com/songquanpeng/one-api/common/config"
+)
+
+var store = make(map[int][]bool)
+var metricSuccessChan = make(chan int, config.MetricSuccessChanSize)
+var metricFailChan = make(chan int, config.MetricFailChanSize)
+
+func consumeSuccess(channelId int) {
+	if len(store[channelId]) > config.MetricQueueSize {
+		store[channelId] = store[channelId][1:]
+	}
+	store[channelId] = append(store[channelId], true)
+}
+
+func consumeFail(channelId int) (bool, float64) {
+	if len(store[channelId]) > config.MetricQueueSize {
+		store[channelId] = store[channelId][1:]
+	}
+	store[channelId] = append(store[channelId], false)
+	successCount := 0
+	for _, success := range store[channelId] {
+		if success {
+			successCount++
+		}
+	}
+	successRate := float64(successCount) / float64(len(store[channelId]))
+	if len(store[channelId]) < config.MetricQueueSize {
+		return false, successRate
+	}
+	if successRate < config.MetricSuccessRateThreshold {
+		store[channelId] = make([]bool, 0)
+		return true, successRate
+	}
+	return false, successRate
+}
+
+func metricSuccessConsumer() {
+	for {
+		select {
+		case channelId := <-metricSuccessChan:
+			consumeSuccess(channelId)
+		}
+	}
+}
+
+func metricFailConsumer() {
+	for {
+		select {
+		case channelId := <-metricFailChan:
+			disable, successRate := consumeFail(channelId)
+			if disable {
+				go MetricDisableChannel(channelId, successRate)
+			}
+		}
+	}
+}
+
+func init() {
+	if config.EnableMetric {
+		go metricSuccessConsumer()
+		go metricFailConsumer()
+	}
+}
+
+func Emit(channelId int, success bool) {
+	if !config.EnableMetric {
+		return
+	}
+	go func() {
+		if success {
+			metricSuccessChan <- channelId
+		} else {
+			metricFailChan <- channelId
+		}
+	}()
+}

pull_request_template.md

@@ -1,9 +1,10 @@
 [//]: # (请按照以下格式关联 issue)
-[//]: # (请在提交 PR 前确认所提交的功能可用，附上截图即可，这将有助于项目维护者 review & merge 该 PR，谢谢)
+[//]: # (请在提交 PR 前确认所提交的功能可用，需要附上截图，谢谢)
 [//]: # (项目维护者一般仅在周末处理 PR，因此如若未能及时回复希望能理解)
 [//]: # (开发者交流群：910657413)
 [//]: # (请在提交 PR 之前删除上面的注释)
 
 close #issue_number
 
-我已确认该 PR 已自测通过，相关截图如下：
+我已确认该 PR 已自测通过，相关截图如下：
+（此处放上测试通过的截图，如果不涉及前端改动或从 UI 上无法看出，请放终端启动成功的截图）

relay/channel/ali/adaptor.go

@@ -32,6 +32,9 @@ func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
 	channel.SetupCommonRequestHeader(c, req, meta)
+	if meta.IsStream {
+		req.Header.Set("Accept", "text/event-stream")
+	}
 	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
 	if meta.IsStream {
 		req.Header.Set("X-DashScope-SSE", "enable")

relay/channel/ali/main.go

@@ -48,6 +48,9 @@ func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
 			MaxTokens:         request.MaxTokens,
 			Temperature:       request.Temperature,
 			TopP:              request.TopP,
+			TopK:              request.TopK,
+			ResultFormat:      "message",
+			Tools:             request.Tools,
 		},
 	}
 }
@@ -117,19 +120,11 @@ func embeddingResponseAli2OpenAI(response *EmbeddingResponse) *openai.EmbeddingR
 }
 
 func responseAli2OpenAI(response *ChatResponse) *openai.TextResponse {
-	choice := openai.TextResponseChoice{
-		Index: 0,
-		Message: model.Message{
-			Role:    "assistant",
-			Content: response.Output.Text,
-		},
-		FinishReason: response.Output.FinishReason,
-	}
 	fullTextResponse := openai.TextResponse{
 		Id:      response.RequestId,
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
-		Choices: []openai.TextResponseChoice{choice},
+		Choices: response.Output.Choices,
 		Usage: model.Usage{
 			PromptTokens:     response.Usage.InputTokens,
 			CompletionTokens: response.Usage.OutputTokens,
@@ -140,10 +135,14 @@ func responseAli2OpenAI(response *ChatResponse) *openai.TextResponse {
 }
 
 func streamResponseAli2OpenAI(aliResponse *ChatResponse) *openai.ChatCompletionsStreamResponse {
+	if len(aliResponse.Output.Choices) == 0 {
+		return nil
+	}
+	aliChoice := aliResponse.Output.Choices[0]
 	var choice openai.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = aliResponse.Output.Text
-	if aliResponse.Output.FinishReason != "null" {
-		finishReason := aliResponse.Output.FinishReason
+	choice.Delta = aliChoice.Message
+	if aliChoice.FinishReason != "null" {
+		finishReason := aliChoice.FinishReason
 		choice.FinishReason = &finishReason
 	}
 	response := openai.ChatCompletionsStreamResponse{
@@ -204,6 +203,9 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 				usage.TotalTokens = aliResponse.Usage.InputTokens + aliResponse.Usage.OutputTokens
 			}
 			response := streamResponseAli2OpenAI(&aliResponse)
+			if response == nil {
+				return true
+			}
 			//response.Choices[0].Delta.Content = strings.TrimPrefix(response.Choices[0].Delta.Content, lastResponseText)
 			//lastResponseText = aliResponse.Output.Text
 			jsonResponse, err := json.Marshal(response)
@@ -226,6 +228,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 }
 
 func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	ctx := c.Request.Context()
 	var aliResponse ChatResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -235,6 +238,7 @@ func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *
 	if err != nil {
 		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
+	logger.Debugf(ctx, "response body: %s\n", responseBody)
 	err = json.Unmarshal(responseBody, &aliResponse)
 	if err != nil {
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil

relay/channel/ali/model.go

@@ -1,5 +1,10 @@
 package ali
 
+import (
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
 type Message struct {
 	Content string `json:"content"`
 	Role    string `json:"role"`
@@ -11,13 +16,15 @@ type Input struct {
 }
 
 type Parameters struct {
-	TopP              float64 `json:"top_p,omitempty"`
-	TopK              int     `json:"top_k,omitempty"`
-	Seed              uint64  `json:"seed,omitempty"`
-	EnableSearch      bool    `json:"enable_search,omitempty"`
-	IncrementalOutput bool    `json:"incremental_output,omitempty"`
-	MaxTokens         int     `json:"max_tokens,omitempty"`
-	Temperature       float64 `json:"temperature,omitempty"`
+	TopP              float64      `json:"top_p,omitempty"`
+	TopK              int          `json:"top_k,omitempty"`
+	Seed              uint64       `json:"seed,omitempty"`
+	EnableSearch      bool         `json:"enable_search,omitempty"`
+	IncrementalOutput bool         `json:"incremental_output,omitempty"`
+	MaxTokens         int          `json:"max_tokens,omitempty"`
+	Temperature       float64      `json:"temperature,omitempty"`
+	ResultFormat      string       `json:"result_format,omitempty"`
+	Tools             []model.Tool `json:"tools,omitempty"`
 }
 
 type ChatRequest struct {
@@ -62,8 +69,9 @@ type Usage struct {
 }
 
 type Output struct {
-	Text         string `json:"text"`
-	FinishReason string `json:"finish_reason"`
+	//Text         string                      `json:"text"`
+	//FinishReason string                      `json:"finish_reason"`
+	Choices []openai.TextResponseChoice `json:"choices"`
 }
 
 type ChatResponse struct {

relay/channel/anthropic/adaptor.go

@@ -5,7 +5,6 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/relay/channel"
-	"github.com/songquanpeng/one-api/relay/channel/openai"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/util"
 	"io"
@@ -20,7 +19,7 @@ func (a *Adaptor) Init(meta *util.RelayMeta) {
 }
 
 func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
-	return fmt.Sprintf("%s/v1/complete", meta.BaseURL), nil
+	return fmt.Sprintf("%s/v1/messages", meta.BaseURL), nil
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
@@ -31,6 +30,7 @@ func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *ut
 		anthropicVersion = "2023-06-01"
 	}
 	req.Header.Set("anthropic-version", anthropicVersion)
+	req.Header.Set("anthropic-beta", "messages-2023-12-15")
 	return nil
 }
 
@@ -47,9 +47,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
 	if meta.IsStream {
-		var responseText string
-		err, responseText = StreamHandler(c, resp)
-		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+		err, usage = StreamHandler(c, resp)
 	} else {
 		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
 	}
@@ -61,5 +59,5 @@ func (a *Adaptor) GetModelList() []string {
 }
 
 func (a *Adaptor) GetChannelName() string {
-	return "authropic"
+	return "anthropic"
 }

relay/channel/anthropic/constants.go

@@ -1,5 +1,8 @@
 package anthropic
 
 var ModelList = []string{
-	"claude-instant-1", "claude-2", "claude-2.0", "claude-2.1",
+	"claude-instant-1.2", "claude-2.0", "claude-2.1",
+	"claude-3-haiku-20240307",
+	"claude-3-sonnet-20240229",
+	"claude-3-opus-20240229",
 }

relay/channel/anthropic/main.go

@@ -7,6 +7,7 @@ import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/image"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/channel/openai"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -15,73 +16,136 @@ import (
 	"strings"
 )
 
-func stopReasonClaude2OpenAI(reason string) string {
-	switch reason {
+func stopReasonClaude2OpenAI(reason *string) string {
+	if reason == nil {
+		return ""
+	}
+	switch *reason {
+	case "end_turn":
+		return "stop"
 	case "stop_sequence":
 		return "stop"
 	case "max_tokens":
 		return "length"
 	default:
-		return reason
+		return *reason
 	}
 }
 
 func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
 	claudeRequest := Request{
-		Model:             textRequest.Model,
-		Prompt:            "",
-		MaxTokensToSample: textRequest.MaxTokens,
-		StopSequences:     nil,
-		Temperature:       textRequest.Temperature,
-		TopP:              textRequest.TopP,
-		Stream:            textRequest.Stream,
+		Model:       textRequest.Model,
+		MaxTokens:   textRequest.MaxTokens,
+		Temperature: textRequest.Temperature,
+		TopP:        textRequest.TopP,
+		TopK:        textRequest.TopK,
+		Stream:      textRequest.Stream,
 	}
-	if claudeRequest.MaxTokensToSample == 0 {
-		claudeRequest.MaxTokensToSample = 1000000
+	if claudeRequest.MaxTokens == 0 {
+		claudeRequest.MaxTokens = 4096
+	}
+	// legacy model name mapping
+	if claudeRequest.Model == "claude-instant-1" {
+		claudeRequest.Model = "claude-instant-1.1"
+	} else if claudeRequest.Model == "claude-2" {
+		claudeRequest.Model = "claude-2.1"
 	}
-	prompt := ""
 	for _, message := range textRequest.Messages {
-		if message.Role == "user" {
-			prompt += fmt.Sprintf("\n\nHuman: %s", message.Content)
-		} else if message.Role == "assistant" {
-			prompt += fmt.Sprintf("\n\nAssistant: %s", message.Content)
-		} else if message.Role == "system" {
-			if prompt == "" {
-				prompt = message.StringContent()
-			}
+		if message.Role == "system" && claudeRequest.System == "" {
+			claudeRequest.System = message.StringContent()
+			continue
 		}
+		claudeMessage := Message{
+			Role: message.Role,
+		}
+		var content Content
+		if message.IsStringContent() {
+			content.Type = "text"
+			content.Text = message.StringContent()
+			claudeMessage.Content = append(claudeMessage.Content, content)
+			claudeRequest.Messages = append(claudeRequest.Messages, claudeMessage)
+			continue
+		}
+		var contents []Content
+		openaiContent := message.ParseContent()
+		for _, part := range openaiContent {
+			var content Content
+			if part.Type == model.ContentTypeText {
+				content.Type = "text"
+				content.Text = part.Text
+			} else if part.Type == model.ContentTypeImageURL {
+				content.Type = "image"
+				content.Source = &ImageSource{
+					Type: "base64",
+				}
+				mimeType, data, _ := image.GetImageFromUrl(part.ImageURL.Url)
+				content.Source.MediaType = mimeType
+				content.Source.Data = data
+			}
+			contents = append(contents, content)
+		}
+		claudeMessage.Content = contents
+		claudeRequest.Messages = append(claudeRequest.Messages, claudeMessage)
 	}
-	prompt += "\n\nAssistant:"
-	claudeRequest.Prompt = prompt
 	return &claudeRequest
 }
 
-func streamResponseClaude2OpenAI(claudeResponse *Response) *openai.ChatCompletionsStreamResponse {
+// https://docs.anthropic.com/claude/reference/messages-streaming
+func streamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCompletionsStreamResponse, *Response) {
+	var response *Response
+	var responseText string
+	var stopReason string
+	switch claudeResponse.Type {
+	case "message_start":
+		return nil, claudeResponse.Message
+	case "content_block_start":
+		if claudeResponse.ContentBlock != nil {
+			responseText = claudeResponse.ContentBlock.Text
+		}
+	case "content_block_delta":
+		if claudeResponse.Delta != nil {
+			responseText = claudeResponse.Delta.Text
+		}
+	case "message_delta":
+		if claudeResponse.Usage != nil {
+			response = &Response{
+				Usage: *claudeResponse.Usage,
+			}
+		}
+		if claudeResponse.Delta != nil && claudeResponse.Delta.StopReason != nil {
+			stopReason = *claudeResponse.Delta.StopReason
+		}
+	}
 	var choice openai.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = claudeResponse.Completion
-	finishReason := stopReasonClaude2OpenAI(claudeResponse.StopReason)
+	choice.Delta.Content = responseText
+	choice.Delta.Role = "assistant"
+	finishReason := stopReasonClaude2OpenAI(&stopReason)
 	if finishReason != "null" {
 		choice.FinishReason = &finishReason
 	}
-	var response openai.ChatCompletionsStreamResponse
-	response.Object = "chat.completion.chunk"
-	response.Model = claudeResponse.Model
-	response.Choices = []openai.ChatCompletionsStreamResponseChoice{choice}
-	return &response
+	var openaiResponse openai.ChatCompletionsStreamResponse
+	openaiResponse.Object = "chat.completion.chunk"
+	openaiResponse.Choices = []openai.ChatCompletionsStreamResponseChoice{choice}
+	return &openaiResponse, response
 }
 
 func responseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
+	var responseText string
+	if len(claudeResponse.Content) > 0 {
+		responseText = claudeResponse.Content[0].Text
+	}
 	choice := openai.TextResponseChoice{
 		Index: 0,
 		Message: model.Message{
 			Role:    "assistant",
-			Content: strings.TrimPrefix(claudeResponse.Completion, " "),
+			Content: responseText,
 			Name:    nil,
 		},
 		FinishReason: stopReasonClaude2OpenAI(claudeResponse.StopReason),
 	}
 	fullTextResponse := openai.TextResponse{
-		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
+		Id:      fmt.Sprintf("chatcmpl-%s", claudeResponse.Id),
+		Model:   claudeResponse.Model,
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
 		Choices: []openai.TextResponseChoice{choice},
@@ -89,17 +153,15 @@ func responseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
 	return &fullTextResponse
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, string) {
-	responseText := ""
-	responseId := fmt.Sprintf("chatcmpl-%s", helper.GetUUID())
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	createdTime := helper.GetTimestamp()
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
 			return 0, nil, nil
 		}
-		if i := strings.Index(string(data), "\r\n\r\n"); i >= 0 {
-			return i + 4, data[0:i], nil
+		if i := strings.Index(string(data), "\n"); i >= 0 {
+			return i + 1, data[0:i], nil
 		}
 		if atEOF {
 			return len(data), data, nil
@@ -111,29 +173,45 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 	go func() {
 		for scanner.Scan() {
 			data := scanner.Text()
-			if !strings.HasPrefix(data, "event: completion") {
+			if len(data) < 6 {
 				continue
 			}
-			data = strings.TrimPrefix(data, "event: completion\r\ndata: ")
+			if !strings.HasPrefix(data, "data: ") {
+				continue
+			}
+			data = strings.TrimPrefix(data, "data: ")
 			dataChan <- data
 		}
 		stopChan <- true
 	}()
 	common.SetEventStreamHeaders(c)
+	var usage model.Usage
+	var modelName string
+	var id string
 	c.Stream(func(w io.Writer) bool {
 		select {
 		case data := <-dataChan:
 			// some implementations may add \r at the end of data
 			data = strings.TrimSuffix(data, "\r")
-			var claudeResponse Response
+			var claudeResponse StreamResponse
 			err := json.Unmarshal([]byte(data), &claudeResponse)
 			if err != nil {
 				logger.SysError("error unmarshalling stream response: " + err.Error())
 				return true
 			}
-			responseText += claudeResponse.Completion
-			response := streamResponseClaude2OpenAI(&claudeResponse)
-			response.Id = responseId
+			response, meta := streamResponseClaude2OpenAI(&claudeResponse)
+			if meta != nil {
+				usage.PromptTokens += meta.Usage.InputTokens
+				usage.CompletionTokens += meta.Usage.OutputTokens
+				modelName = meta.Model
+				id = fmt.Sprintf("chatcmpl-%s", meta.Id)
+				return true
+			}
+			if response == nil {
+				return true
+			}
+			response.Id = id
+			response.Model = modelName
 			response.Created = createdTime
 			jsonStr, err := json.Marshal(response)
 			if err != nil {
@@ -147,11 +225,8 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 			return false
 		}
 	})
-	err := resp.Body.Close()
-	if err != nil {
-		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
-	}
-	return nil, responseText
+	_ = resp.Body.Close()
+	return nil, &usage
 }
 
 func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
@@ -181,11 +256,10 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 	}
 	fullTextResponse := responseClaude2OpenAI(&claudeResponse)
 	fullTextResponse.Model = modelName
-	completionTokens := openai.CountTokenText(claudeResponse.Completion, modelName)
 	usage := model.Usage{
-		PromptTokens:     promptTokens,
-		CompletionTokens: completionTokens,
-		TotalTokens:      promptTokens + completionTokens,
+		PromptTokens:     claudeResponse.Usage.InputTokens,
+		CompletionTokens: claudeResponse.Usage.OutputTokens,
+		TotalTokens:      claudeResponse.Usage.InputTokens + claudeResponse.Usage.OutputTokens,
 	}
 	fullTextResponse.Usage = usage
 	jsonResponse, err := json.Marshal(fullTextResponse)

relay/channel/anthropic/model.go

@@ -1,19 +1,44 @@
 package anthropic
 
+// https://docs.anthropic.com/claude/reference/messages_post
+
 type Metadata struct {
 	UserId string `json:"user_id"`
 }
 
+type ImageSource struct {
+	Type      string `json:"type"`
+	MediaType string `json:"media_type"`
+	Data      string `json:"data"`
+}
+
+type Content struct {
+	Type   string       `json:"type"`
+	Text   string       `json:"text,omitempty"`
+	Source *ImageSource `json:"source,omitempty"`
+}
+
+type Message struct {
+	Role    string    `json:"role"`
+	Content []Content `json:"content"`
+}
+
 type Request struct {
-	Model             string   `json:"model"`
-	Prompt            string   `json:"prompt"`
-	MaxTokensToSample int      `json:"max_tokens_to_sample"`
-	StopSequences     []string `json:"stop_sequences,omitempty"`
-	Temperature       float64  `json:"temperature,omitempty"`
-	TopP              float64  `json:"top_p,omitempty"`
-	TopK              int      `json:"top_k,omitempty"`
+	Model         string    `json:"model"`
+	Messages      []Message `json:"messages"`
+	System        string    `json:"system,omitempty"`
+	MaxTokens     int       `json:"max_tokens,omitempty"`
+	StopSequences []string  `json:"stop_sequences,omitempty"`
+	Stream        bool      `json:"stream,omitempty"`
+	Temperature   float64   `json:"temperature,omitempty"`
+	TopP          float64   `json:"top_p,omitempty"`
+	TopK          int       `json:"top_k,omitempty"`
 	//Metadata    `json:"metadata,omitempty"`
-	Stream bool `json:"stream,omitempty"`
+}
+
+type Usage struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
 }
 
 type Error struct {
@@ -22,8 +47,29 @@ type Error struct {
 }
 
 type Response struct {
-	Completion string `json:"completion"`
-	StopReason string `json:"stop_reason"`
-	Model      string `json:"model"`
-	Error      Error  `json:"error"`
+	Id           string    `json:"id"`
+	Type         string    `json:"type"`
+	Role         string    `json:"role"`
+	Content      []Content `json:"content"`
+	Model        string    `json:"model"`
+	StopReason   *string   `json:"stop_reason"`
+	StopSequence *string   `json:"stop_sequence"`
+	Usage        Usage     `json:"usage"`
+	Error        Error     `json:"error"`
+}
+
+type Delta struct {
+	Type         string  `json:"type"`
+	Text         string  `json:"text"`
+	StopReason   *string `json:"stop_reason"`
+	StopSequence *string `json:"stop_sequence"`
+}
+
+type StreamResponse struct {
+	Type         string    `json:"type"`
+	Message      *Response `json:"message"`
+	Index        int       `json:"index"`
+	ContentBlock *Content  `json:"content_block"`
+	Delta        *Delta    `json:"delta"`
+	Usage        *Usage    `json:"usage"`
 }

relay/channel/baidu/adaptor.go

@@ -2,13 +2,16 @@ package baidu
 
 import (
 	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/relay/channel"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/util"
-	"io"
-	"net/http"
 )
 
 type Adaptor struct {
@@ -20,25 +23,55 @@ func (a *Adaptor) Init(meta *util.RelayMeta) {
 
 func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
 	// https://cloud.baidu.com/doc/WENXINWORKSHOP/s/clntwmv7t
-	var fullRequestURL string
-	switch meta.ActualModelName {
-	case "ERNIE-Bot-4":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro"
-	case "ERNIE-Bot-8K":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie_bot_8k"
-	case "ERNIE-Bot":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions"
-	case "ERNIE-Speed":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie_speed"
-	case "ERNIE-Bot-turbo":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/eb-instant"
-	case "BLOOMZ-7B":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/bloomz_7b1"
-	case "Embedding-V1":
-		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/embeddings/embedding-v1"
-	default:
-               fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/" + meta.ActualModelName
+	suffix := "chat/"
+	if strings.HasPrefix(meta.ActualModelName, "Embedding") {
+		suffix = "embeddings/"
 	}
+	if strings.HasPrefix(meta.ActualModelName, "bge-large") {
+		suffix = "embeddings/"
+	}
+	if strings.HasPrefix(meta.ActualModelName, "tao-8k") {
+		suffix = "embeddings/"
+	}
+	switch meta.ActualModelName {
+	case "ERNIE-4.0":
+		suffix += "completions_pro"
+	case "ERNIE-Bot-4":
+		suffix += "completions_pro"
+	case "ERNIE-Bot":
+		suffix += "completions"
+	case "ERNIE-Bot-turbo":
+		suffix += "eb-instant"
+	case "ERNIE-Speed":
+		suffix += "ernie_speed"
+	case "ERNIE-Bot-8K":
+		suffix += "ernie_bot_8k"
+	case "ERNIE-4.0-8K":
+		suffix += "completions_pro"
+	case "ERNIE-3.5-8K":
+		suffix += "completions"
+	case "ERNIE-Speed-8K":
+		suffix += "ernie_speed"
+	case "ERNIE-Speed-128K":
+		suffix += "ernie-speed-128k"
+	case "ERNIE-Lite-8K":
+		suffix += "ernie-lite-8k"
+	case "ERNIE-Tiny-8K":
+		suffix += "ernie-tiny-8k"
+	case "BLOOMZ-7B":
+		suffix += "bloomz_7b1"
+	case "Embedding-V1":
+		suffix += "embedding-v1"
+	case "bge-large-zh":
+		suffix += "bge_large_zh"
+	case "bge-large-en":
+		suffix += "bge_large_en"
+	case "tao-8k":
+		suffix += "tao_8k"
+	default:
+		suffix += strings.ToLower(meta.ActualModelName)
+	}
+	fullRequestURL := fmt.Sprintf("%s/rpc/2.0/ai_custom/v1/wenxinworkshop/%s", meta.BaseURL, suffix)
 	var accessToken string
 	var err error
 	if accessToken, err = GetAccessToken(meta.APIKey); err != nil {

relay/channel/baidu/constants.go

@@ -1,10 +1,20 @@
 package baidu
 
 var ModelList = []string{
-	"ERNIE-Bot-4",
-	"ERNIE-Bot-8K",
-	"ERNIE-Bot",
-	"ERNIE-Speed",
-	"ERNIE-Bot-turbo",
+	"ERNIE-4.0-8K",
+	"ERNIE-Bot-8K-0922",
+	"ERNIE-3.5-8K",
+	"ERNIE-Lite-8K-0922",
+	"ERNIE-Speed-8K",
+	"ERNIE-3.5-4K-0205",
+	"ERNIE-3.5-8K-0205",
+	"ERNIE-3.5-8K-1222",
+	"ERNIE-Lite-8K",
+	"ERNIE-Speed-128K",
+	"ERNIE-Tiny-8K",
+	"BLOOMZ-7B",
 	"Embedding-V1",
+	"bge-large-zh",
+	"bge-large-en",
+	"tao-8k",
 }

relay/channel/baidu/main.go

@@ -32,9 +32,16 @@ type Message struct {
 }
 
 type ChatRequest struct {
-	Messages []Message `json:"messages"`
-	Stream   bool      `json:"stream"`
-	UserId   string    `json:"user_id,omitempty"`
+	Messages        []Message `json:"messages"`
+	Temperature     float64   `json:"temperature,omitempty"`
+	TopP            float64   `json:"top_p,omitempty"`
+	PenaltyScore    float64   `json:"penalty_score,omitempty"`
+	Stream          bool      `json:"stream,omitempty"`
+	System          string    `json:"system,omitempty"`
+	DisableSearch   bool      `json:"disable_search,omitempty"`
+	EnableCitation  bool      `json:"enable_citation,omitempty"`
+	MaxOutputTokens int       `json:"max_output_tokens,omitempty"`
+	UserId          string    `json:"user_id,omitempty"`
 }
 
 type Error struct {
@@ -45,28 +52,28 @@ type Error struct {
 var baiduTokenStore sync.Map
 
 func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
-	messages := make([]Message, 0, len(request.Messages))
+	baiduRequest := ChatRequest{
+		Messages:        make([]Message, 0, len(request.Messages)),
+		Temperature:     request.Temperature,
+		TopP:            request.TopP,
+		PenaltyScore:    request.FrequencyPenalty,
+		Stream:          request.Stream,
+		DisableSearch:   false,
+		EnableCitation:  false,
+		MaxOutputTokens: request.MaxTokens,
+		UserId:          request.User,
+	}
 	for _, message := range request.Messages {
 		if message.Role == "system" {
-			messages = append(messages, Message{
-				Role:    "user",
-				Content: message.StringContent(),
-			})
-			messages = append(messages, Message{
-				Role:    "assistant",
-				Content: "Okay",
-			})
+			baiduRequest.System = message.StringContent()
 		} else {
-			messages = append(messages, Message{
+			baiduRequest.Messages = append(baiduRequest.Messages, Message{
 				Role:    message.Role,
 				Content: message.StringContent(),
 			})
 		}
 	}
-	return &ChatRequest{
-		Messages: messages,
-		Stream:   request.Stream,
-	}
+	return &baiduRequest
 }
 
 func responseBaidu2OpenAI(response *ChatResponse) *openai.TextResponse {

relay/channel/gemini/constants.go

@@ -1,6 +1,8 @@
 package gemini
 
+// https://ai.google.dev/models/gemini
+
 var ModelList = []string{
-	"gemini-pro", "gemini-1.0-pro-001",
+	"gemini-pro", "gemini-1.0-pro-001", "gemini-1.5-pro",
 	"gemini-pro-vision", "gemini-1.0-pro-vision-001",
 }

relay/channel/groq/constants.go

@@ -0,0 +1,10 @@
+package groq
+
+// https://console.groq.com/docs/models
+
+var ModelList = []string{
+	"gemma-7b-it",
+	"llama2-7b-2048",
+	"llama2-70b-4096",
+	"mixtral-8x7b-32768",
+}

relay/channel/lingyiwanwu/constants.go

@@ -0,0 +1,9 @@
+package lingyiwanwu
+
+// https://platform.lingyiwanwu.com/docs
+
+var ModelList = []string{
+	"yi-34b-chat-0205",
+	"yi-34b-chat-200k",
+	"yi-vl-plus",
+}

relay/channel/ollama/adaptor.go

@@ -0,0 +1,75 @@
+package ollama
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+)
+
+type Adaptor struct {
+}
+
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	// https://github.com/ollama/ollama/blob/main/docs/api.md
+	fullRequestURL := fmt.Sprintf("%s/api/chat", meta.BaseURL)
+	if meta.Mode == constant.RelayModeEmbeddings {
+		fullRequestURL = fmt.Sprintf("%s/api/embeddings", meta.BaseURL)
+	}
+	return fullRequestURL, nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	switch relayMode {
+	case constant.RelayModeEmbeddings:
+		ollamaEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return ollamaEmbeddingRequest, nil
+	default:
+		return ConvertRequest(*request), nil
+	}
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp)
+	} else {
+		switch meta.Mode {
+		case constant.RelayModeEmbeddings:
+			err, usage = EmbeddingHandler(c, resp)
+		default:
+			err, usage = Handler(c, resp)
+		}
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "ollama"
+}

relay/channel/ollama/constants.go

@@ -0,0 +1,5 @@
+package ollama
+
+var ModelList = []string{
+	"qwen:0.5b-chat",
+}

relay/channel/ollama/main.go

@@ -0,0 +1,237 @@
+package ollama
+
+import (
+	"bufio"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
+	ollamaRequest := ChatRequest{
+		Model: request.Model,
+		Options: &Options{
+			Seed:             int(request.Seed),
+			Temperature:      request.Temperature,
+			TopP:             request.TopP,
+			FrequencyPenalty: request.FrequencyPenalty,
+			PresencePenalty:  request.PresencePenalty,
+		},
+		Stream: request.Stream,
+	}
+	for _, message := range request.Messages {
+		ollamaRequest.Messages = append(ollamaRequest.Messages, Message{
+			Role:    message.Role,
+			Content: message.StringContent(),
+		})
+	}
+	return &ollamaRequest
+}
+
+func responseOllama2OpenAI(response *ChatResponse) *openai.TextResponse {
+	choice := openai.TextResponseChoice{
+		Index: 0,
+		Message: model.Message{
+			Role:    response.Message.Role,
+			Content: response.Message.Content,
+		},
+	}
+	if response.Done {
+		choice.FinishReason = "stop"
+	}
+	fullTextResponse := openai.TextResponse{
+		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
+		Object:  "chat.completion",
+		Created: helper.GetTimestamp(),
+		Choices: []openai.TextResponseChoice{choice},
+		Usage: model.Usage{
+			PromptTokens:     response.PromptEvalCount,
+			CompletionTokens: response.EvalCount,
+			TotalTokens:      response.PromptEvalCount + response.EvalCount,
+		},
+	}
+	return &fullTextResponse
+}
+
+func streamResponseOllama2OpenAI(ollamaResponse *ChatResponse) *openai.ChatCompletionsStreamResponse {
+	var choice openai.ChatCompletionsStreamResponseChoice
+	choice.Delta.Role = ollamaResponse.Message.Role
+	choice.Delta.Content = ollamaResponse.Message.Content
+	if ollamaResponse.Done {
+		choice.FinishReason = &constant.StopFinishReason
+	}
+	response := openai.ChatCompletionsStreamResponse{
+		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
+		Object:  "chat.completion.chunk",
+		Created: helper.GetTimestamp(),
+		Model:   ollamaResponse.Model,
+		Choices: []openai.ChatCompletionsStreamResponseChoice{choice},
+	}
+	return &response
+}
+
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var usage model.Usage
+	scanner := bufio.NewScanner(resp.Body)
+	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
+		if atEOF && len(data) == 0 {
+			return 0, nil, nil
+		}
+		if i := strings.Index(string(data), "}\n"); i >= 0 {
+			return i + 2, data[0:i], nil
+		}
+		if atEOF {
+			return len(data), data, nil
+		}
+		return 0, nil, nil
+	})
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
+	go func() {
+		for scanner.Scan() {
+			data := strings.TrimPrefix(scanner.Text(), "}")
+			dataChan <- data + "}"
+		}
+		stopChan <- true
+	}()
+	common.SetEventStreamHeaders(c)
+	c.Stream(func(w io.Writer) bool {
+		select {
+		case data := <-dataChan:
+			var ollamaResponse ChatResponse
+			err := json.Unmarshal([]byte(data), &ollamaResponse)
+			if err != nil {
+				logger.SysError("error unmarshalling stream response: " + err.Error())
+				return true
+			}
+			if ollamaResponse.EvalCount != 0 {
+				usage.PromptTokens = ollamaResponse.PromptEvalCount
+				usage.CompletionTokens = ollamaResponse.EvalCount
+				usage.TotalTokens = ollamaResponse.PromptEvalCount + ollamaResponse.EvalCount
+			}
+			response := streamResponseOllama2OpenAI(&ollamaResponse)
+			jsonResponse, err := json.Marshal(response)
+			if err != nil {
+				logger.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonResponse)})
+			return true
+		case <-stopChan:
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+	})
+	err := resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	return nil, &usage
+}
+
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingRequest {
+	return &EmbeddingRequest{
+		Model:  request.Model,
+		Prompt: strings.Join(request.ParseInput(), " "),
+	}
+}
+
+func EmbeddingHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var ollamaResponse EmbeddingResponse
+	err := json.NewDecoder(resp.Body).Decode(&ollamaResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+
+	if ollamaResponse.Error != "" {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: ollamaResponse.Error,
+				Type:    "ollama_error",
+				Param:   "",
+				Code:    "ollama_error",
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+
+	fullTextResponse := embeddingResponseOllama2OpenAI(&ollamaResponse)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &fullTextResponse.Usage
+}
+
+func embeddingResponseOllama2OpenAI(response *EmbeddingResponse) *openai.EmbeddingResponse {
+	openAIEmbeddingResponse := openai.EmbeddingResponse{
+		Object: "list",
+		Data:   make([]openai.EmbeddingResponseItem, 0, 1),
+		Model:  "text-embedding-v1",
+		Usage:  model.Usage{TotalTokens: 0},
+	}
+
+	openAIEmbeddingResponse.Data = append(openAIEmbeddingResponse.Data, openai.EmbeddingResponseItem{
+		Object:    `embedding`,
+		Index:     0,
+		Embedding: response.Embedding,
+	})
+	return &openAIEmbeddingResponse
+}
+
+func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	ctx := context.TODO()
+	var ollamaResponse ChatResponse
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	logger.Debugf(ctx, "ollama response: %s", string(responseBody))
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = json.Unmarshal(responseBody, &ollamaResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	if ollamaResponse.Error != "" {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: ollamaResponse.Error,
+				Type:    "ollama_error",
+				Param:   "",
+				Code:    "ollama_error",
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+	fullTextResponse := responseOllama2OpenAI(&ollamaResponse)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &fullTextResponse.Usage
+}

relay/channel/ollama/model.go

@@ -0,0 +1,47 @@
+package ollama
+
+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"`
+}
+
+type Message struct {
+	Role    string   `json:"role,omitempty"`
+	Content string   `json:"content,omitempty"`
+	Images  []string `json:"images,omitempty"`
+}
+
+type ChatRequest struct {
+	Model    string    `json:"model,omitempty"`
+	Messages []Message `json:"messages,omitempty"`
+	Stream   bool      `json:"stream"`
+	Options  *Options  `json:"options,omitempty"`
+}
+
+type ChatResponse struct {
+	Model           string  `json:"model,omitempty"`
+	CreatedAt       string  `json:"created_at,omitempty"`
+	Message         Message `json:"message,omitempty"`
+	Response        string  `json:"response,omitempty"` // for stream response
+	Done            bool    `json:"done,omitempty"`
+	TotalDuration   int     `json:"total_duration,omitempty"`
+	LoadDuration    int     `json:"load_duration,omitempty"`
+	PromptEvalCount int     `json:"prompt_eval_count,omitempty"`
+	EvalCount       int     `json:"eval_count,omitempty"`
+	EvalDuration    int     `json:"eval_duration,omitempty"`
+	Error           string  `json:"error,omitempty"`
+}
+
+type EmbeddingRequest struct {
+	Model  string `json:"model"`
+	Prompt string `json:"prompt"`
+}
+
+type EmbeddingResponse struct {
+	Error     string    `json:"error,omitempty"`
+	Embedding []float64 `json:"embedding,omitempty"`
+}

relay/channel/openai/adaptor.go

@@ -6,11 +6,7 @@ import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/relay/channel"
-	"github.com/songquanpeng/one-api/relay/channel/ai360"
-	"github.com/songquanpeng/one-api/relay/channel/baichuan"
 	"github.com/songquanpeng/one-api/relay/channel/minimax"
-	"github.com/songquanpeng/one-api/relay/channel/mistral"
-	"github.com/songquanpeng/one-api/relay/channel/moonshot"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/util"
 	"io"
@@ -35,11 +31,8 @@ func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
 		task := strings.TrimPrefix(requestURL, "/v1/")
 		model_ := meta.ActualModelName
 		model_ = strings.Replace(model_, ".", "", -1)
-		// https://github.com/songquanpeng/one-api/issues/67
-		model_ = strings.TrimSuffix(model_, "-0301")
-		model_ = strings.TrimSuffix(model_, "-0314")
-		model_ = strings.TrimSuffix(model_, "-0613")
-
+		//https://github.com/songquanpeng/one-api/issues/1191
+		// {your endpoint}/openai/deployments/{your azure_model}/chat/completions?api-version={api_version}
 		requestURL = fmt.Sprintf("/openai/deployments/%s/%s", model_, task)
 		return util.GetFullRequestURL(meta.BaseURL, requestURL, meta.ChannelType), nil
 	case common.ChannelTypeMinimax:
@@ -77,8 +70,10 @@ func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
 	if meta.IsStream {
 		var responseText string
-		err, responseText, _ = StreamHandler(c, resp, meta.Mode)
-		usage = ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+		err, responseText, usage = StreamHandler(c, resp, meta.Mode)
+		if usage == nil {
+			usage = ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+		}
 	} else {
 		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
 	}
@@ -86,37 +81,11 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.Rel
 }
 
 func (a *Adaptor) GetModelList() []string {
-	switch a.ChannelType {
-	case common.ChannelType360:
-		return ai360.ModelList
-	case common.ChannelTypeMoonshot:
-		return moonshot.ModelList
-	case common.ChannelTypeBaichuan:
-		return baichuan.ModelList
-	case common.ChannelTypeMinimax:
-		return minimax.ModelList
-	case common.ChannelTypeMistral:
-		return mistral.ModelList
-	default:
-		return ModelList
-	}
+	_, modelList := GetCompatibleChannelMeta(a.ChannelType)
+	return modelList
 }
 
 func (a *Adaptor) GetChannelName() string {
-	switch a.ChannelType {
-	case common.ChannelTypeAzure:
-		return "azure"
-	case common.ChannelType360:
-		return "360"
-	case common.ChannelTypeMoonshot:
-		return "moonshot"
-	case common.ChannelTypeBaichuan:
-		return "baichuan"
-	case common.ChannelTypeMinimax:
-		return "minimax"
-	case common.ChannelTypeMistral:
-		return "mistralai"
-	default:
-		return "openai"
-	}
+	channelName, _ := GetCompatibleChannelMeta(a.ChannelType)
+	return channelName
 }

relay/channel/openai/compatible.go

@@ -0,0 +1,50 @@
+package openai
+
+import (
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/relay/channel/ai360"
+	"github.com/songquanpeng/one-api/relay/channel/baichuan"
+	"github.com/songquanpeng/one-api/relay/channel/groq"
+	"github.com/songquanpeng/one-api/relay/channel/lingyiwanwu"
+	"github.com/songquanpeng/one-api/relay/channel/minimax"
+	"github.com/songquanpeng/one-api/relay/channel/mistral"
+	"github.com/songquanpeng/one-api/relay/channel/moonshot"
+	"github.com/songquanpeng/one-api/relay/channel/stepfun"
+)
+
+var CompatibleChannels = []int{
+	common.ChannelTypeAzure,
+	common.ChannelType360,
+	common.ChannelTypeMoonshot,
+	common.ChannelTypeBaichuan,
+	common.ChannelTypeMinimax,
+	common.ChannelTypeMistral,
+	common.ChannelTypeGroq,
+	common.ChannelTypeLingYiWanWu,
+	common.ChannelTypeStepFun,
+}
+
+func GetCompatibleChannelMeta(channelType int) (string, []string) {
+	switch channelType {
+	case common.ChannelTypeAzure:
+		return "azure", ModelList
+	case common.ChannelType360:
+		return "360", ai360.ModelList
+	case common.ChannelTypeMoonshot:
+		return "moonshot", moonshot.ModelList
+	case common.ChannelTypeBaichuan:
+		return "baichuan", baichuan.ModelList
+	case common.ChannelTypeMinimax:
+		return "minimax", minimax.ModelList
+	case common.ChannelTypeMistral:
+		return "mistralai", mistral.ModelList
+	case common.ChannelTypeGroq:
+		return "groq", groq.ModelList
+	case common.ChannelTypeLingYiWanWu:
+		return "lingyiwanwu", lingyiwanwu.ModelList
+	case common.ChannelTypeStepFun:
+		return "stepfun", stepfun.ModelList
+	default:
+		return "openai", ModelList
+	}
+}

relay/channel/openai/main.go

@@ -6,6 +6,7 @@ import (
 	"encoding/json"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/conv"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -53,7 +54,7 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.E
 						continue // just ignore the error
 					}
 					for _, choice := range streamResponse.Choices {
-						responseText += choice.Delta.Content
+						responseText += conv.AsString(choice.Delta.Content)
 					}
 					if streamResponse.Usage != nil {
 						usage = streamResponse.Usage

relay/channel/openai/model.go

@@ -118,12 +118,9 @@ type ImageResponse struct {
 }
 
 type ChatCompletionsStreamResponseChoice struct {
-	Index int `json:"index"`
-	Delta struct {
-		Content string `json:"content"`
-		Role    string `json:"role,omitempty"`
-	} `json:"delta"`
-	FinishReason *string `json:"finish_reason,omitempty"`
+	Index        int           `json:"index"`
+	Delta        model.Message `json:"delta"`
+	FinishReason *string       `json:"finish_reason,omitempty"`
 }
 
 type ChatCompletionsStreamResponse struct {

relay/channel/stepfun/constants.go

@@ -0,0 +1,7 @@
+package stepfun
+
+var ModelList = []string{
+	"step-1-32k",
+	"step-1v-32k",
+	"step-1-200k",
+}

relay/channel/tencent/main.go

@@ -10,6 +10,7 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/conv"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/channel/openai"
@@ -28,17 +29,6 @@ func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
 	messages := make([]Message, 0, len(request.Messages))
 	for i := 0; i < len(request.Messages); i++ {
 		message := request.Messages[i]
-		if message.Role == "system" {
-			messages = append(messages, Message{
-				Role:    "user",
-				Content: message.StringContent(),
-			})
-			messages = append(messages, Message{
-				Role:    "assistant",
-				Content: "Okay",
-			})
-			continue
-		}
 		messages = append(messages, Message{
 			Content: message.StringContent(),
 			Role:    message.Role,
@@ -140,7 +130,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 			}
 			response := streamResponseTencent2OpenAI(&TencentResponse)
 			if len(response.Choices) != 0 {
-				responseText += response.Choices[0].Delta.Content
+				responseText += conv.AsString(response.Choices[0].Delta.Content)
 			}
 			jsonResponse, err := json.Marshal(response)
 			if err != nil {

relay/channel/xunfei/main.go

@@ -26,22 +26,15 @@ import (
 
 func requestOpenAI2Xunfei(request model.GeneralOpenAIRequest, xunfeiAppId string, domain string) *ChatRequest {
 	messages := make([]Message, 0, len(request.Messages))
+	var lastToolCalls []model.Tool
 	for _, message := range request.Messages {
-		if message.Role == "system" {
-			messages = append(messages, Message{
-				Role:    "user",
-				Content: message.StringContent(),
-			})
-			messages = append(messages, Message{
-				Role:    "assistant",
-				Content: "Okay",
-			})
-		} else {
-			messages = append(messages, Message{
-				Role:    message.Role,
-				Content: message.StringContent(),
-			})
+		if message.ToolCalls != nil {
+			lastToolCalls = message.ToolCalls
 		}
+		messages = append(messages, Message{
+			Role:    message.Role,
+			Content: message.StringContent(),
+		})
 	}
 	xunfeiRequest := ChatRequest{}
 	xunfeiRequest.Header.AppId = xunfeiAppId
@@ -50,9 +43,33 @@ func requestOpenAI2Xunfei(request model.GeneralOpenAIRequest, xunfeiAppId string
 	xunfeiRequest.Parameter.Chat.TopK = request.N
 	xunfeiRequest.Parameter.Chat.MaxTokens = request.MaxTokens
 	xunfeiRequest.Payload.Message.Text = messages
+	if len(lastToolCalls) != 0 {
+		for _, toolCall := range lastToolCalls {
+			xunfeiRequest.Payload.Functions.Text = append(xunfeiRequest.Payload.Functions.Text, toolCall.Function)
+		}
+	}
+
 	return &xunfeiRequest
 }
 
+func getToolCalls(response *ChatResponse) []model.Tool {
+	var toolCalls []model.Tool
+	if len(response.Payload.Choices.Text) == 0 {
+		return toolCalls
+	}
+	item := response.Payload.Choices.Text[0]
+	if item.FunctionCall == nil {
+		return toolCalls
+	}
+	toolCall := model.Tool{
+		Id:       fmt.Sprintf("call_%s", helper.GetUUID()),
+		Type:     "function",
+		Function: *item.FunctionCall,
+	}
+	toolCalls = append(toolCalls, toolCall)
+	return toolCalls
+}
+
 func responseXunfei2OpenAI(response *ChatResponse) *openai.TextResponse {
 	if len(response.Payload.Choices.Text) == 0 {
 		response.Payload.Choices.Text = []ChatResponseTextItem{
@@ -64,8 +81,9 @@ func responseXunfei2OpenAI(response *ChatResponse) *openai.TextResponse {
 	choice := openai.TextResponseChoice{
 		Index: 0,
 		Message: model.Message{
-			Role:    "assistant",
-			Content: response.Payload.Choices.Text[0].Content,
+			Role:      "assistant",
+			Content:   response.Payload.Choices.Text[0].Content,
+			ToolCalls: getToolCalls(response),
 		},
 		FinishReason: constant.StopFinishReason,
 	}
@@ -89,6 +107,7 @@ func streamResponseXunfei2OpenAI(xunfeiResponse *ChatResponse) *openai.ChatCompl
 	}
 	var choice openai.ChatCompletionsStreamResponseChoice
 	choice.Delta.Content = xunfeiResponse.Payload.Choices.Text[0].Content
+	choice.Delta.ToolCalls = getToolCalls(xunfeiResponse)
 	if xunfeiResponse.Payload.Choices.Status == 2 {
 		choice.FinishReason = &constant.StopFinishReason
 	}
@@ -132,7 +151,7 @@ func StreamHandler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId
 	domain, authUrl := getXunfeiAuthUrl(c, apiKey, apiSecret, textRequest.Model)
 	dataChan, stopChan, err := xunfeiMakeRequest(textRequest, domain, authUrl, appId)
 	if err != nil {
-		return openai.ErrorWrapper(err, "make xunfei request err", http.StatusInternalServerError), nil
+		return openai.ErrorWrapper(err, "xunfei_request_failed", http.StatusInternalServerError), nil
 	}
 	common.SetEventStreamHeaders(c)
 	var usage model.Usage
@@ -162,7 +181,7 @@ func Handler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId strin
 	domain, authUrl := getXunfeiAuthUrl(c, apiKey, apiSecret, textRequest.Model)
 	dataChan, stopChan, err := xunfeiMakeRequest(textRequest, domain, authUrl, appId)
 	if err != nil {
-		return openai.ErrorWrapper(err, "make xunfei request err", http.StatusInternalServerError), nil
+		return openai.ErrorWrapper(err, "xunfei_request_failed", http.StatusInternalServerError), nil
 	}
 	var usage model.Usage
 	var content string
@@ -182,11 +201,7 @@ func Handler(c *gin.Context, textRequest model.GeneralOpenAIRequest, appId strin
 		}
 	}
 	if len(xunfeiResponse.Payload.Choices.Text) == 0 {
-		xunfeiResponse.Payload.Choices.Text = []ChatResponseTextItem{
-			{
-				Content: "",
-			},
-		}
+		return openai.ErrorWrapper(err, "xunfei_empty_response_detected", http.StatusInternalServerError), nil
 	}
 	xunfeiResponse.Payload.Choices.Text[0].Content = content
 
@@ -213,15 +228,21 @@ func xunfeiMakeRequest(textRequest model.GeneralOpenAIRequest, domain, authUrl,
 	if err != nil {
 		return nil, nil, err
 	}
+	_, msg, err := conn.ReadMessage()
+	if err != nil {
+		return nil, nil, err
+	}
 
 	dataChan := make(chan ChatResponse)
 	stopChan := make(chan bool)
 	go func() {
 		for {
-			_, msg, err := conn.ReadMessage()
-			if err != nil {
-				logger.SysError("error reading stream response: " + err.Error())
-				break
+			if msg == nil {
+				_, msg, err = conn.ReadMessage()
+				if err != nil {
+					logger.SysError("error reading stream response: " + err.Error())
+					break
+				}
 			}
 			var response ChatResponse
 			err = json.Unmarshal(msg, &response)
@@ -229,6 +250,7 @@ func xunfeiMakeRequest(textRequest model.GeneralOpenAIRequest, domain, authUrl,
 				logger.SysError("error unmarshalling stream response: " + err.Error())
 				break
 			}
+			msg = nil
 			dataChan <- response
 			if response.Payload.Choices.Status == 2 {
 				err := conn.Close()

relay/channel/xunfei/model.go

@@ -26,13 +26,18 @@ type ChatRequest struct {
 		Message struct {
 			Text []Message `json:"text"`
 		} `json:"message"`
+		Functions struct {
+			Text []model.Function `json:"text,omitempty"`
+		} `json:"functions,omitempty"`
 	} `json:"payload"`
 }
 
 type ChatResponseTextItem struct {
-	Content string `json:"content"`
-	Role    string `json:"role"`
-	Index   int    `json:"index"`
+	Content      string          `json:"content"`
+	Role         string          `json:"role"`
+	Index        int             `json:"index"`
+	ContentType  string          `json:"content_type"`
+	FunctionCall *model.Function `json:"function_call"`
 }
 
 type ChatResponse struct {

relay/channel/zhipu/adaptor.go

@@ -6,9 +6,11 @@ import (
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/relay/channel"
 	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/util"
 	"io"
+	"math"
 	"net/http"
 	"strings"
 )
@@ -34,6 +36,9 @@ func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
 	if a.APIVersion == "v4" {
 		return fmt.Sprintf("%s/api/paas/v4/chat/completions", meta.BaseURL), nil
 	}
+	if meta.Mode == constant.RelayModeEmbeddings {
+		return fmt.Sprintf("%s/api/paas/v4/embeddings", meta.BaseURL), nil
+	}
 	method := "invoke"
 	if meta.IsStream {
 		method = "sse-invoke"
@@ -52,14 +57,24 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	if request.TopP >= 1 {
-		request.TopP = 0.99
+	switch relayMode {
+	case constant.RelayModeEmbeddings:
+		baiduEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return baiduEmbeddingRequest, nil
+	default:
+		// TopP (0.0, 1.0)
+		request.TopP = math.Min(0.99, request.TopP)
+		request.TopP = math.Max(0.01, request.TopP)
+
+		// Temperature (0.0, 1.0)
+		request.Temperature = math.Min(0.99, request.Temperature)
+		request.Temperature = math.Max(0.01, request.Temperature)
+		a.SetVersionByModeName(request.Model)
+		if a.APIVersion == "v4" {
+			return request, nil
+		}
+		return ConvertRequest(*request), nil
 	}
-	a.SetVersionByModeName(request.Model)
-	if a.APIVersion == "v4" {
-		return request, nil
-	}
-	return ConvertRequest(*request), nil
 }
 
 func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
@@ -79,14 +94,26 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.Rel
 	if a.APIVersion == "v4" {
 		return a.DoResponseV4(c, resp, meta)
 	}
+
 	if meta.IsStream {
 		err, usage = StreamHandler(c, resp)
 	} else {
-		err, usage = Handler(c, resp)
+		if meta.Mode == constant.RelayModeEmbeddings {
+			err, usage = EmbeddingsHandler(c, resp)
+		} else {
+			err, usage = Handler(c, resp)
+		}
 	}
 	return
 }
 
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingRequest {
+	return &EmbeddingRequest{
+		Model: "embedding-2",
+		Input: request.Input.(string),
+	}
+}
+
 func (a *Adaptor) GetModelList() []string {
 	return ModelList
 }

relay/channel/zhipu/constants.go

@@ -2,5 +2,5 @@ package zhipu
 
 var ModelList = []string{
 	"chatglm_turbo", "chatglm_pro", "chatglm_std", "chatglm_lite",
-	"glm-4", "glm-4v", "glm-3-turbo",
+	"glm-4", "glm-4v", "glm-3-turbo", "embedding-2",
 }

relay/channel/zhipu/main.go

@@ -76,21 +76,10 @@ func GetToken(apikey string) string {
 func ConvertRequest(request model.GeneralOpenAIRequest) *Request {
 	messages := make([]Message, 0, len(request.Messages))
 	for _, message := range request.Messages {
-		if message.Role == "system" {
-			messages = append(messages, Message{
-				Role:    "system",
-				Content: message.StringContent(),
-			})
-			messages = append(messages, Message{
-				Role:    "user",
-				Content: "Okay",
-			})
-		} else {
-			messages = append(messages, Message{
-				Role:    message.Role,
-				Content: message.StringContent(),
-			})
-		}
+		messages = append(messages, Message{
+			Role:    message.Role,
+			Content: message.StringContent(),
+		})
 	}
 	return &Request{
 		Prompt:      messages,
@@ -265,3 +254,50 @@ func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *
 	_, err = c.Writer.Write(jsonResponse)
 	return nil, &fullTextResponse.Usage
 }
+
+func EmbeddingsHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var zhipuResponse EmbeddingRespone
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = json.Unmarshal(responseBody, &zhipuResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	fullTextResponse := embeddingResponseZhipu2OpenAI(&zhipuResponse)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &fullTextResponse.Usage
+}
+
+func embeddingResponseZhipu2OpenAI(response *EmbeddingRespone) *openai.EmbeddingResponse {
+	openAIEmbeddingResponse := openai.EmbeddingResponse{
+		Object: "list",
+		Data:   make([]openai.EmbeddingResponseItem, 0, len(response.Embeddings)),
+		Model:  response.Model,
+		Usage: model.Usage{
+			PromptTokens:     response.PromptTokens,
+			CompletionTokens: response.CompletionTokens,
+			TotalTokens:      response.Usage.TotalTokens,
+		},
+	}
+
+	for _, item := range response.Embeddings {
+		openAIEmbeddingResponse.Data = append(openAIEmbeddingResponse.Data, openai.EmbeddingResponseItem{
+			Object:    `embedding`,
+			Index:     item.Index,
+			Embedding: item.Embedding,
+		})
+	}
+	return &openAIEmbeddingResponse
+}

relay/channel/zhipu/model.go

@@ -44,3 +44,21 @@ type tokenData struct {
 	Token      string
 	ExpiryTime time.Time
 }
+
+type EmbeddingRequest struct {
+	Model string `json:"model"`
+	Input string `json:"input"`
+}
+
+type EmbeddingRespone struct {
+	Model       string          `json:"model"`
+	Object      string          `json:"object"`
+	Embeddings  []EmbeddingData `json:"data"`
+	model.Usage `json:"usage"`
+}
+
+type EmbeddingData struct {
+	Index     int       `json:"index"`
+	Object    string    `json:"object"`
+	Embedding []float64 `json:"embedding"`
+}

relay/constant/api_type.go

@@ -15,6 +15,7 @@ const (
 	APITypeAIProxyLibrary
 	APITypeTencent
 	APITypeGemini
+	APITypeOllama
 
 	APITypeDummy // this one is only for count, do not add any channel after this
 )
@@ -40,6 +41,8 @@ func ChannelType2APIType(channelType int) int {
 		apiType = APITypeTencent
 	case common.ChannelTypeGemini:
 		apiType = APITypeGemini
+	case common.ChannelTypeOllama:
+		apiType = APITypeOllama
 	}
 	return apiType
 }

relay/controller/audio.go

@@ -22,6 +22,7 @@ import (
 )
 
 func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatusCode {
+	ctx := c.Request.Context()
 	audioModel := "whisper-1"
 
 	tokenId := c.GetInt("token_id")
@@ -49,16 +50,16 @@ func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 	modelRatio := common.GetModelRatio(audioModel)
 	groupRatio := common.GetGroupRatio(group)
 	ratio := modelRatio * groupRatio
-	var quota int
-	var preConsumedQuota int
+	var quota int64
+	var preConsumedQuota int64
 	switch relayMode {
 	case constant.RelayModeAudioSpeech:
-		preConsumedQuota = int(float64(len(ttsRequest.Input)) * ratio)
+		preConsumedQuota = int64(float64(len(ttsRequest.Input)) * ratio)
 		quota = preConsumedQuota
 	default:
-		preConsumedQuota = int(float64(config.PreConsumedQuota) * ratio)
+		preConsumedQuota = int64(float64(config.PreConsumedQuota) * ratio)
 	}
-	userQuota, err := model.CacheGetUserQuota(userId)
+	userQuota, err := model.CacheGetUserQuota(ctx, userId)
 	if err != nil {
 		return openai.ErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
@@ -82,6 +83,24 @@ func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 			return openai.ErrorWrapper(err, "pre_consume_token_quota_failed", http.StatusForbidden)
 		}
 	}
+	succeed := false
+	defer func() {
+		if succeed {
+			return
+		}
+		if preConsumedQuota > 0 {
+			// we need to roll back the pre-consumed quota
+			defer func(ctx context.Context) {
+				go func() {
+					// negative means add quota back for token & user
+					err := model.PostConsumeTokenQuota(tokenId, -preConsumedQuota)
+					if err != nil {
+						logger.Error(ctx, fmt.Sprintf("error rollback pre-consumed quota: %s", err.Error()))
+					}
+				}()
+			}(c.Request.Context())
+		}
+	}()
 
 	// map model name
 	modelMapping := c.GetString("model_mapping")
@@ -103,10 +122,15 @@ func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 	}
 
 	fullRequestURL := util.GetFullRequestURL(baseURL, requestURL, channelType)
-	if relayMode == constant.RelayModeAudioTranscription && channelType == common.ChannelTypeAzure {
-		// https://learn.microsoft.com/en-us/azure/ai-services/openai/whisper-quickstart?tabs=command-line#rest-api
+	if channelType == common.ChannelTypeAzure {
 		apiVersion := util.GetAzureAPIVersion(c)
-		fullRequestURL = fmt.Sprintf("%s/openai/deployments/%s/audio/transcriptions?api-version=%s", baseURL, audioModel, apiVersion)
+		if relayMode == constant.RelayModeAudioTranscription {
+			// https://learn.microsoft.com/en-us/azure/ai-services/openai/whisper-quickstart?tabs=command-line#rest-api
+			fullRequestURL = fmt.Sprintf("%s/openai/deployments/%s/audio/transcriptions?api-version=%s", baseURL, audioModel, apiVersion)
+		} else if relayMode == constant.RelayModeAudioSpeech {
+			// https://learn.microsoft.com/en-us/azure/ai-services/openai/text-to-speech-quickstart?tabs=command-line#rest-api
+			fullRequestURL = fmt.Sprintf("%s/openai/deployments/%s/audio/speech?api-version=%s", baseURL, audioModel, apiVersion)
+		}
 	}
 
 	requestBody := &bytes.Buffer{}
@@ -122,7 +146,7 @@ func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 		return openai.ErrorWrapper(err, "new_request_failed", http.StatusInternalServerError)
 	}
 
-	if relayMode == constant.RelayModeAudioTranscription && channelType == common.ChannelTypeAzure {
+	if (relayMode == constant.RelayModeAudioTranscription || relayMode == constant.RelayModeAudioSpeech) && channelType == common.ChannelTypeAzure {
 		// https://learn.microsoft.com/en-us/azure/ai-services/openai/whisper-quickstart?tabs=command-line#rest-api
 		apiKey := c.Request.Header.Get("Authorization")
 		apiKey = strings.TrimPrefix(apiKey, "Bearer ")
@@ -183,24 +207,13 @@ func RelayAudioHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 		if err != nil {
 			return openai.ErrorWrapper(err, "get_text_from_body_err", http.StatusInternalServerError)
 		}
-		quota = openai.CountTokenText(text, audioModel)
+		quota = int64(openai.CountTokenText(text, audioModel))
 		resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
 	}
 	if resp.StatusCode != http.StatusOK {
-		if preConsumedQuota > 0 {
-			// we need to roll back the pre-consumed quota
-			defer func(ctx context.Context) {
-				go func() {
-					// negative means add quota back for token & user
-					err := model.PostConsumeTokenQuota(tokenId, -preConsumedQuota)
-					if err != nil {
-						logger.Error(ctx, fmt.Sprintf("error rollback pre-consumed quota: %s", err.Error()))
-					}
-				}()
-			}(c.Request.Context())
-		}
 		return util.RelayErrorHandler(resp)
 	}
+	succeed = true
 	quotaDelta := quota - preConsumedQuota
 	defer func(ctx context.Context) {
 		go util.PostConsumeQuota(ctx, tokenId, quotaDelta, quota, userId, channelId, modelRatio, groupRatio, audioModel, tokenName)

relay/controller/helper.go

@@ -107,18 +107,18 @@ func getPromptTokens(textRequest *relaymodel.GeneralOpenAIRequest, relayMode int
 	return 0
 }
 
-func getPreConsumedQuota(textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64) int {
+func getPreConsumedQuota(textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64) int64 {
 	preConsumedTokens := config.PreConsumedQuota
 	if textRequest.MaxTokens != 0 {
-		preConsumedTokens = promptTokens + textRequest.MaxTokens
+		preConsumedTokens = int64(promptTokens) + int64(textRequest.MaxTokens)
 	}
-	return int(float64(preConsumedTokens) * ratio)
+	return int64(float64(preConsumedTokens) * ratio)
 }
 
-func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64, meta *util.RelayMeta) (int, *relaymodel.ErrorWithStatusCode) {
+func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64, meta *util.RelayMeta) (int64, *relaymodel.ErrorWithStatusCode) {
 	preConsumedQuota := getPreConsumedQuota(textRequest, promptTokens, ratio)
 
-	userQuota, err := model.CacheGetUserQuota(meta.UserId)
+	userQuota, err := model.CacheGetUserQuota(ctx, meta.UserId)
 	if err != nil {
 		return preConsumedQuota, openai.ErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
@@ -144,16 +144,16 @@ func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIR
 	return preConsumedQuota, nil
 }
 
-func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.RelayMeta, textRequest *relaymodel.GeneralOpenAIRequest, ratio float64, preConsumedQuota int, modelRatio float64, groupRatio float64) {
+func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.RelayMeta, textRequest *relaymodel.GeneralOpenAIRequest, ratio float64, preConsumedQuota int64, modelRatio float64, groupRatio float64) {
 	if usage == nil {
 		logger.Error(ctx, "usage is nil, which is unexpected")
 		return
 	}
-	quota := 0
+	var quota int64
 	completionRatio := common.GetCompletionRatio(textRequest.Model)
 	promptTokens := usage.PromptTokens
 	completionTokens := usage.CompletionTokens
-	quota = int(math.Ceil((float64(promptTokens) + float64(completionTokens)*completionRatio) * ratio))
+	quota = int64(math.Ceil((float64(promptTokens) + float64(completionTokens)*completionRatio) * ratio))
 	if ratio != 0 && quota <= 0 {
 		quota = 1
 	}
@@ -168,7 +168,7 @@ func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.R
 	if err != nil {
 		logger.Error(ctx, "error consuming token remain quota: "+err.Error())
 	}
-	err = model.CacheUpdateUserQuota(meta.UserId)
+	err = model.CacheUpdateUserQuota(ctx, meta.UserId)
 	if err != nil {
 		logger.Error(ctx, "error update user quota cache: "+err.Error())
 	}

relay/controller/image.go

@@ -61,7 +61,7 @@ func RelayImageHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 	if meta.ChannelType == common.ChannelTypeAzure {
 		// https://learn.microsoft.com/en-us/azure/ai-services/openai/dall-e-quickstart?tabs=dalle3%2Ccommand-line&pivots=rest-api
 		apiVersion := util.GetAzureAPIVersion(c)
-		// https://{resource_name}.openai.azure.com/openai/deployments/dall-e-3/images/generations?api-version=2023-06-01-preview
+		// https://{resource_name}.openai.azure.com/openai/deployments/dall-e-3/images/generations?api-version=2024-03-01-preview
 		fullRequestURL = fmt.Sprintf("%s/openai/deployments/%s/images/generations?api-version=%s", meta.BaseURL, imageRequest.Model, apiVersion)
 	}
 
@@ -79,9 +79,9 @@ func RelayImageHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 	modelRatio := common.GetModelRatio(imageRequest.Model)
 	groupRatio := common.GetGroupRatio(meta.Group)
 	ratio := modelRatio * groupRatio
-	userQuota, err := model.CacheGetUserQuota(meta.UserId)
+	userQuota, err := model.CacheGetUserQuota(ctx, meta.UserId)
 
-	quota := int(ratio*imageCostRatio*1000) * imageRequest.N
+	quota := int64(ratio*imageCostRatio*1000) * int64(imageRequest.N)
 
 	if userQuota-quota < 0 {
 		return openai.ErrorWrapper(errors.New("user quota is not enough"), "insufficient_user_quota", http.StatusForbidden)
@@ -125,7 +125,7 @@ func RelayImageHelper(c *gin.Context, relayMode int) *relaymodel.ErrorWithStatus
 		if err != nil {
 			logger.SysError("error consuming token remain quota: " + err.Error())
 		}
-		err = model.CacheUpdateUserQuota(meta.UserId)
+		err = model.CacheUpdateUserQuota(ctx, meta.UserId)
 		if err != nil {
 			logger.SysError("error update user quota cache: " + err.Error())
 		}

relay/controller/text.go

@@ -74,6 +74,7 @@ func RelayTextHelper(c *gin.Context) *model.ErrorWithStatusCode {
 		if err != nil {
 			return openai.ErrorWrapper(err, "json_marshal_failed", http.StatusInternalServerError)
 		}
+		logger.Debugf(ctx, "converted request: \n%s", string(jsonData))
 		requestBody = bytes.NewBuffer(jsonData)
 	}
 
@@ -83,11 +84,12 @@ func RelayTextHelper(c *gin.Context) *model.ErrorWithStatusCode {
 		logger.Errorf(ctx, "DoRequest failed: %s", err.Error())
 		return openai.ErrorWrapper(err, "do_request_failed", http.StatusInternalServerError)
 	}
-	meta.IsStream = meta.IsStream || strings.HasPrefix(resp.Header.Get("Content-Type"), "text/event-stream")
-	if resp.StatusCode != http.StatusOK {
+	errorHappened := (resp.StatusCode != http.StatusOK) || (meta.IsStream && resp.Header.Get("Content-Type") == "application/json")
+	if errorHappened {
 		util.ReturnPreConsumedQuota(ctx, preConsumedQuota, meta.TokenId)
 		return util.RelayErrorHandler(resp)
 	}
+	meta.IsStream = meta.IsStream || strings.HasPrefix(resp.Header.Get("Content-Type"), "text/event-stream")
 
 	// do response
 	usage, respErr := adaptor.DoResponse(c, resp, meta)

relay/helper/main.go

@@ -7,6 +7,7 @@ import (
 	"github.com/songquanpeng/one-api/relay/channel/anthropic"
 	"github.com/songquanpeng/one-api/relay/channel/baidu"
 	"github.com/songquanpeng/one-api/relay/channel/gemini"
+	"github.com/songquanpeng/one-api/relay/channel/ollama"
 	"github.com/songquanpeng/one-api/relay/channel/openai"
 	"github.com/songquanpeng/one-api/relay/channel/palm"
 	"github.com/songquanpeng/one-api/relay/channel/tencent"
@@ -37,6 +38,8 @@ func GetAdaptor(apiType int) channel.Adaptor {
 		return &xunfei.Adaptor{}
 	case constant.APITypeZhipu:
 		return &zhipu.Adaptor{}
+	case constant.APITypeOllama:
+		return &ollama.Adaptor{}
 	}
 	return nil
 }

relay/model/general.go

@@ -5,25 +5,29 @@ type ResponseFormat struct {
 }
 
 type GeneralOpenAIRequest struct {
-	Model            string          `json:"model,omitempty"`
 	Messages         []Message       `json:"messages,omitempty"`
-	Prompt           any             `json:"prompt,omitempty"`
-	Stream           bool            `json:"stream,omitempty"`
-	MaxTokens        int             `json:"max_tokens,omitempty"`
-	Temperature      float64         `json:"temperature,omitempty"`
-	TopP             float64         `json:"top_p,omitempty"`
-	N                int             `json:"n,omitempty"`
-	Input            any             `json:"input,omitempty"`
-	Instruction      string          `json:"instruction,omitempty"`
-	Size             string          `json:"size,omitempty"`
-	Functions        any             `json:"functions,omitempty"`
+	Model            string          `json:"model,omitempty"`
 	FrequencyPenalty float64         `json:"frequency_penalty,omitempty"`
+	MaxTokens        int             `json:"max_tokens,omitempty"`
+	N                int             `json:"n,omitempty"`
 	PresencePenalty  float64         `json:"presence_penalty,omitempty"`
 	ResponseFormat   *ResponseFormat `json:"response_format,omitempty"`
 	Seed             float64         `json:"seed,omitempty"`
-	Tools            any             `json:"tools,omitempty"`
+	Stream           bool            `json:"stream,omitempty"`
+	Temperature      float64         `json:"temperature,omitempty"`
+	TopP             float64         `json:"top_p,omitempty"`
+	TopK             int             `json:"top_k,omitempty"`
+	Tools            []Tool          `json:"tools,omitempty"`
 	ToolChoice       any             `json:"tool_choice,omitempty"`
+	FunctionCall     any             `json:"function_call,omitempty"`
+	Functions        any             `json:"functions,omitempty"`
 	User             string          `json:"user,omitempty"`
+	Prompt           any             `json:"prompt,omitempty"`
+	Input            any             `json:"input,omitempty"`
+	EncodingFormat   string          `json:"encoding_format,omitempty"`
+	Dimensions       int             `json:"dimensions,omitempty"`
+	Instruction      string          `json:"instruction,omitempty"`
+	Size             string          `json:"size,omitempty"`
 }
 
 func (r GeneralOpenAIRequest) ParseInput() []string {

relay/model/message.go

@@ -1,9 +1,10 @@
 package model
 
 type Message struct {
-	Role    string  `json:"role"`
-	Content any     `json:"content"`
-	Name    *string `json:"name,omitempty"`
+	Role      string  `json:"role,omitempty"`
+	Content   any     `json:"content,omitempty"`
+	Name      *string `json:"name,omitempty"`
+	ToolCalls []Tool  `json:"tool_calls,omitempty"`
 }
 
 func (m Message) IsStringContent() bool {

relay/model/tool.go

@@ -0,0 +1,14 @@
+package model
+
+type Tool struct {
+	Id       string   `json:"id,omitempty"`
+	Type     string   `json:"type"`
+	Function Function `json:"function"`
+}
+
+type Function struct {
+	Description string `json:"description,omitempty"`
+	Name        string `json:"name"`
+	Parameters  any    `json:"parameters,omitempty"` // request
+	Arguments   any    `json:"arguments,omitempty"`  // response
+}