feat: 支持自定义特殊模型补全倍率

add-mj-use-time

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

@@ -4,7 +4,6 @@ on:
   push:
     tags:
       - '*'
-      - '!*-alpha*'
   workflow_dispatch:
     inputs:
       name:

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

@@ -4,7 +4,6 @@ on:
   push:
     tags:
       - '*'
-      - '!*-alpha*'
   workflow_dispatch:
     inputs:
       name:

Dockerfile

@@ -5,7 +5,7 @@ COPY web/package.json .
 RUN npm install
 COPY ./web .
 COPY ./VERSION .
-RUN DISABLE_ESLINT_PLUGIN='true' REACT_APP_VERSION=$(cat VERSION) npm run build
+RUN DISABLE_ESLINT_PLUGIN='true' VITE_REACT_APP_VERSION=$(cat VERSION) npm run build
 
 FROM golang AS builder2
 
@@ -17,7 +17,7 @@ WORKDIR /build
 ADD go.mod go.sum ./
 RUN go mod download
 COPY . .
-COPY --from=builder /build/build ./web/build
+COPY --from=builder /build/dist ./web/dist
 RUN go build -ldflags "-s -w -X 'one-api/common.Version=$(cat VERSION)' -extldflags '-static'" -o one-api
 
 FROM alpine

README.md

@@ -55,9 +55,20 @@
 3. Anthropic Claude 3 (claude-3-opus-20240229, claude-3-sonnet-20240229)
 4. [Ollama](https://github.com/ollama/ollama?tab=readme-ov-file)，添加渠道时，密钥可以随便填写，默认的请求地址是[http://localhost:11434](http://localhost:11434)，如果需要修改请在渠道中修改
 5. [Midjourney-Proxy(Plus)](https://github.com/novicezk/midjourney-proxy)接口，[对接文档](Midjourney.md)
+6. [零一万物](https://platform.lingyiwanwu.com/)
 
 您可以在渠道中添加自定义模型gpt-4-gizmo-*，此模型并非OpenAI官方模型，而是第三方模型，使用官方key无法调用。
 
+## 渠道重试
+渠道重试功能已经实现，可以在`设置->运营设置->通用设置`设置重试次数，建议开启缓存功能。  
+如果开启了重试功能，第一次重试使用同优先级，第二次重试使用下一个优先级，以此类推。  
+### 缓存设置方法
+1. `REDIS_CONN_STRING`：设置之后将使用 Redis 作为缓存使用。
+    + 例子：`REDIS_CONN_STRING=redis://default:redispw@localhost:49153`
+2. `MEMORY_CACHE_ENABLED`：启用内存缓存（如果设置了`REDIS_CONN_STRING`，则无需手动设置），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`
+
+
 ## 部署
 ### 基于 Docker 进行部署
 ```shell

common/constants.go

@@ -9,15 +9,6 @@ import (
 	"github.com/google/uuid"
 )
 
-// Pay Settings
-
-var PayAddress = ""
-var CustomCallbackAddress = ""
-var EpayId = ""
-var EpayKey = ""
-var Price = 7.3
-var MinTopUp = 1
-
 var StartTime = time.Now().Unix() // unit: second
 var Version = "v0.0.0"            // this hard coding will be replaced automatically when building, no need to manually change
 var SystemName = "New API"
@@ -55,7 +46,8 @@ var TelegramOAuthEnabled = false
 var TurnstileCheckEnabled = false
 var RegisterEnabled = true
 
-var EmailDomainRestrictionEnabled = false
+var EmailDomainRestrictionEnabled = false // 是否启用邮箱域名限制
+var EmailAliasRestrictionEnabled = false  // 是否启用邮箱别名限制
 var EmailDomainWhitelist = []string{
 	"gmail.com",
 	"163.com",
@@ -75,6 +67,7 @@ var LogConsumeEnabled = true
 
 var SMTPServer = ""
 var SMTPPort = 587
+var SMTPSSLEnabled = false
 var SMTPAccount = ""
 var SMTPFrom = ""
 var SMTPToken = ""
@@ -110,7 +103,7 @@ var IsMasterNode = os.Getenv("NODE_TYPE") != "slave"
 var requestInterval, _ = strconv.Atoi(os.Getenv("POLLING_INTERVAL"))
 var RequestInterval = time.Duration(requestInterval) * time.Second
 
-var SyncFrequency = GetOrDefault("SYNC_FREQUENCY", 10*60) // unit is second
+var SyncFrequency = GetOrDefault("SYNC_FREQUENCY", 60) // unit is second
 
 var BatchUpdateEnabled = false
 var BatchUpdateInterval = GetOrDefault("BATCH_UPDATE_INTERVAL", 5)
@@ -212,6 +205,11 @@ const (
 	ChannelTypeMoonshot       = 25
 	ChannelTypeZhipu_v4       = 26
 	ChannelTypePerplexity     = 27
+	ChannelTypeLingYiWanWu    = 31
+	ChannelTypeAws            = 33
+	ChannelTypeCohere         = 34
+
+	ChannelTypeDummy // this one is only for count, do not add any channel after this
 )
 
 var ChannelBaseURLs = []string{
@@ -243,4 +241,11 @@ var ChannelBaseURLs = []string{
 	"https://api.moonshot.cn",                   //25
 	"https://open.bigmodel.cn",                  //26
 	"https://api.perplexity.ai",                 //27
+	"",                                          //28
+	"",                                          //29
+	"",                                          //30
+	"https://api.lingyiwanwu.com",               //31
+	"",                                          //32
+	"",                                          //33
+	"https://api.cohere.ai",                     //34
 }

common/email.go

@@ -24,7 +24,7 @@ func SendEmail(subject string, receiver string, content string) error {
 	addr := fmt.Sprintf("%s:%d", SMTPServer, SMTPPort)
 	to := strings.Split(receiver, ";")
 	var err error
-	if SMTPPort == 465 {
+	if SMTPPort == 465 || SMTPSSLEnabled {
 		tlsConfig := &tls.Config{
 			InsecureSkipVerify: true,
 			ServerName:         SMTPServer,

common/gin.go

@@ -5,18 +5,37 @@ import (
 	"encoding/json"
 	"github.com/gin-gonic/gin"
 	"io"
+	"strings"
 )
 
-func UnmarshalBodyReusable(c *gin.Context, v any) error {
+const KeyRequestBody = "key_request_body"
+
+func GetRequestBody(c *gin.Context) ([]byte, error) {
+	requestBody, _ := c.Get(KeyRequestBody)
+	if requestBody != nil {
+		return requestBody.([]byte), nil
+	}
 	requestBody, err := io.ReadAll(c.Request.Body)
 	if err != nil {
-		return err
+		return nil, err
 	}
-	err = c.Request.Body.Close()
+	_ = c.Request.Body.Close()
+	c.Set(KeyRequestBody, requestBody)
+	return requestBody.([]byte), nil
+}
+
+func UnmarshalBodyReusable(c *gin.Context, v any) error {
+	requestBody, err := GetRequestBody(c)
 	if err != nil {
 		return err
 	}
-	err = json.Unmarshal(requestBody, &v)
+	contentType := c.Request.Header.Get("Content-Type")
+	if strings.HasPrefix(contentType, "application/json") {
+		err = json.Unmarshal(requestBody, &v)
+	} else {
+		// skip for now
+		// TODO: someday non json request have variant model, we will need to implementation this
+	}
 	if err != nil {
 		return err
 	}

common/go-channel.go

@@ -16,7 +16,22 @@ func SafeGoroutine(f func()) {
 	}()
 }
 
-func SafeSend(ch chan bool, value bool) (closed bool) {
+func SafeSendBool(ch chan bool, value bool) (closed bool) {
+	defer func() {
+		// Recover from panic if one occured. A panic would mean the channel was closed.
+		if recover() != nil {
+			closed = true
+		}
+	}()
+
+	// This will panic if the channel is closed.
+	ch <- value
+
+	// If the code reaches here, then the channel was not closed.
+	return false
+}
+
+func SafeSendString(ch chan string, value string) (closed bool) {
 	defer func() {
 		// Recover from panic if one occured. A panic would mean the channel was closed.
 		if recover() != nil {

common/logger.go

@@ -2,6 +2,7 @@ package common
 
 import (
 	"context"
+	"encoding/json"
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"io"
@@ -98,3 +99,13 @@ func LogQuota(quota int) string {
 		return fmt.Sprintf("%d 点额度", quota)
 	}
 }
+
+// LogJson 仅供测试使用 only for test
+func LogJson(ctx context.Context, msg string, obj any) {
+	jsonStr, err := json.Marshal(obj)
+	if err != nil {
+		LogError(ctx, fmt.Sprintf("json marshal failed: %s", err.Error()))
+		return
+	}
+	LogInfo(ctx, fmt.Sprintf("%s | %s", msg, string(jsonStr)))
+}

common/model-ratio.go

@@ -3,96 +3,117 @@ package common
 import (
 	"encoding/json"
 	"strings"
-	"time"
 )
 
-// ModelRatio
+// modelRatio
 // https://platform.openai.com/docs/models/model-endpoint-compatibility
 // https://cloud.baidu.com/doc/WENXINWORKSHOP/s/Blfmc9dlf
 // https://openai.com/pricing
 // TODO: when a new api is enabled, check the pricing here
 // 1 === $0.002 / 1K tokens
 // 1 === ￥0.014 / 1k tokens
+
 var DefaultModelRatio = map[string]float64{
 	//"midjourney":                50,
-	"gpt-4-gizmo-*":             15,
-	"gpt-4":                     15,
-	"gpt-4-0314":                15,
-	"gpt-4-0613":                15,
-	"gpt-4-32k":                 30,
-	"gpt-4-32k-0314":            30,
+	"gpt-4-gizmo-*": 15,
+	"gpt-4":         15,
+	//"gpt-4-0314":                   15, //deprecated
+	"gpt-4-0613": 15,
+	"gpt-4-32k":  30,
+	//"gpt-4-32k-0314":               30, //deprecated
 	"gpt-4-32k-0613":            30,
 	"gpt-4-1106-preview":        5,    // $0.01 / 1K tokens
 	"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-4-1106-vision-preview": 5,    // $0.01 / 1K tokens
-	"gpt-3.5-turbo":             0.75, // $0.0015 / 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
-	"gpt-3.5-turbo-16k-0613":    1.5,
-	"gpt-3.5-turbo-instruct":    0.75, // $0.0015 / 1K tokens
-	"gpt-3.5-turbo-1106":        0.5,  // $0.001 / 1K tokens
-	"gpt-3.5-turbo-0125":        0.25,
-	"babbage-002":               0.2, // $0.0004 / 1K tokens
-	"davinci-002":               1,   // $0.002 / 1K tokens
-	"text-ada-001":              0.2,
-	"text-babbage-001":          0.25,
-	"text-curie-001":            1,
-	"text-davinci-002":          10,
-	"text-davinci-003":          10,
-	"text-davinci-edit-001":     10,
-	"code-davinci-edit-001":     10,
-	"whisper-1":                 15,  // $0.006 / minute -> $0.006 / 150 words -> $0.006 / 200 tokens -> $0.03 / 1k tokens
-	"tts-1":                     7.5, // 1k characters -> $0.015
-	"tts-1-1106":                7.5, // 1k characters -> $0.015
-	"tts-1-hd":                  15,  // 1k characters -> $0.03
-	"tts-1-hd-1106":             15,  // 1k characters -> $0.03
-	"davinci":                   10,
-	"curie":                     10,
-	"babbage":                   10,
-	"ada":                       10,
-	"text-embedding-3-small":    0.01,
-	"text-embedding-3-large":    0.065,
-	"text-embedding-ada-002":    0.05,
-	"text-search-ada-doc-001":   10,
-	"text-moderation-stable":    0.1,
-	"text-moderation-latest":    0.1,
-	"dall-e-2":                  8,
-	"dall-e-3":                  16,
-	"claude-instant-1":          0.4,    // $0.8 / 1M tokens
-	"claude-2.0":                4,      // $8 / 1M tokens
-	"claude-2.1":                4,      // $8 / 1M tokens
-	"claude-3-haiku-20240307":   0.125,  // $0.25 / 1M tokens
-	"claude-3-sonnet-20240229":  1.5,    // $3 / 1M tokens
-	"claude-3-opus-20240229":    7.5,    // $15 / 1M tokens
-	"ERNIE-Bot":                 0.8572, // ￥0.012 / 1k tokens
-	"ERNIE-Bot-turbo":           0.5715, // ￥0.008 / 1k tokens
-	"ERNIE-Bot-4":               8.572,  // ￥0.12 / 1k tokens
-	"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
-	"chatglm_turbo":             0.3572, // ￥0.005 / 1k tokens
-	"chatglm_pro":               0.7143, // ￥0.01 / 1k tokens
-	"chatglm_std":               0.3572, // ￥0.005 / 1k tokens
-	"chatglm_lite":              0.1429, // ￥0.002 / 1k tokens
-	"glm-4":                     7.143,  // ￥0.1 / 1k tokens
-	"glm-4v":                    7.143,  // ￥0.1 / 1k tokens
-	"glm-3-turbo":               0.3572,
-	"qwen-turbo":                0.8572, // ￥0.012 / 1k tokens
-	"qwen-plus":                 10,     // ￥0.14 / 1k tokens
-	"text-embedding-v1":         0.05,   // ￥0.0007 / 1k tokens
-	"SparkDesk-v1.1":            1.2858, // ￥0.018 / 1k tokens
-	"SparkDesk-v2.1":            1.2858, // ￥0.018 / 1k tokens
-	"SparkDesk-v3.1":            1.2858, // ￥0.018 / 1k tokens
-	"SparkDesk-v3.5":            1.2858, // ￥0.018 / 1k tokens
-	"360GPT_S2_V9":              0.8572, // ¥0.012 / 1k tokens
-	"embedding-bert-512-v1":     0.0715, // ¥0.001 / 1k tokens
-	"embedding_s1_v1":           0.0715, // ¥0.001 / 1k tokens
-	"semantic_similarity_s1_v1": 0.0715, // ¥0.001 / 1k tokens
-	"hunyuan":                   7.143,  // ¥0.1 / 1k tokens  // https://cloud.tencent.com/document/product/1729/97731#e0e6be58-60c8-469f-bdeb-6c264ce3b4d0
+	"gpt-4-turbo":               5,    // $0.01 / 1K tokens
+	"gpt-4-turbo-2024-04-09":    5,    // $0.01 / 1K tokens
+	"gpt-3.5-turbo":             0.25, // $0.0015 / 1K tokens
+	//"gpt-3.5-turbo-0301":           0.75, //deprecated
+	"gpt-3.5-turbo-0613":           0.75,
+	"gpt-3.5-turbo-16k":            1.5, // $0.003 / 1K tokens
+	"gpt-3.5-turbo-16k-0613":       1.5,
+	"gpt-3.5-turbo-instruct":       0.75, // $0.0015 / 1K tokens
+	"gpt-3.5-turbo-1106":           0.5,  // $0.001 / 1K tokens
+	"gpt-3.5-turbo-0125":           0.25,
+	"babbage-002":                  0.2, // $0.0004 / 1K tokens
+	"davinci-002":                  1,   // $0.002 / 1K tokens
+	"text-ada-001":                 0.2,
+	"text-babbage-001":             0.25,
+	"text-curie-001":               1,
+	"text-davinci-002":             10,
+	"text-davinci-003":             10,
+	"text-davinci-edit-001":        10,
+	"code-davinci-edit-001":        10,
+	"whisper-1":                    15,  // $0.006 / minute -> $0.006 / 150 words -> $0.006 / 200 tokens -> $0.03 / 1k tokens
+	"tts-1":                        7.5, // 1k characters -> $0.015
+	"tts-1-1106":                   7.5, // 1k characters -> $0.015
+	"tts-1-hd":                     15,  // 1k characters -> $0.03
+	"tts-1-hd-1106":                15,  // 1k characters -> $0.03
+	"davinci":                      10,
+	"curie":                        10,
+	"babbage":                      10,
+	"ada":                          10,
+	"text-embedding-3-small":       0.01,
+	"text-embedding-3-large":       0.065,
+	"text-embedding-ada-002":       0.05,
+	"text-search-ada-doc-001":      10,
+	"text-moderation-stable":       0.1,
+	"text-moderation-latest":       0.1,
+	"dall-e-2":                     8,
+	"dall-e-3":                     16,
+	"claude-instant-1":             0.4,    // $0.8 / 1M tokens
+	"claude-2.0":                   4,      // $8 / 1M tokens
+	"claude-2.1":                   4,      // $8 / 1M tokens
+	"claude-3-haiku-20240307":      0.125,  // $0.25 / 1M tokens
+	"claude-3-sonnet-20240229":     1.5,    // $3 / 1M tokens
+	"claude-3-opus-20240229":       7.5,    // $15 / 1M tokens
+	"ERNIE-Bot":                    0.8572, // ￥0.012 / 1k tokens
+	"ERNIE-Bot-turbo":              0.5715, // ￥0.008 / 1k tokens
+	"ERNIE-Bot-4":                  8.572,  // ￥0.12 / 1k tokens
+	"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
+	"gemini-1.0-pro-vision-001":    1,
+	"gemini-1.0-pro-001":           1,
+	"gemini-1.5-pro-latest":        1,
+	"gemini-1.0-pro-latest":        1,
+	"gemini-1.0-pro-vision-latest": 1,
+	"gemini-ultra":                 1,
+	"chatglm_turbo":                0.3572, // ￥0.005 / 1k tokens
+	"chatglm_pro":                  0.7143, // ￥0.01 / 1k tokens
+	"chatglm_std":                  0.3572, // ￥0.005 / 1k tokens
+	"chatglm_lite":                 0.1429, // ￥0.002 / 1k tokens
+	"glm-4":                        7.143,  // ￥0.1 / 1k tokens
+	"glm-4v":                       7.143,  // ￥0.1 / 1k tokens
+	"glm-3-turbo":                  0.3572,
+	"qwen-turbo":                   0.8572, // ￥0.012 / 1k tokens
+	"qwen-plus":                    10,     // ￥0.14 / 1k tokens
+	"text-embedding-v1":            0.05,   // ￥0.0007 / 1k tokens
+	"SparkDesk-v1.1":               1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v2.1":               1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v3.1":               1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v3.5":               1.2858, // ￥0.018 / 1k tokens
+	"360GPT_S2_V9":                 0.8572, // ¥0.012 / 1k tokens
+	"embedding-bert-512-v1":        0.0715, // ¥0.001 / 1k tokens
+	"embedding_s1_v1":              0.0715, // ¥0.001 / 1k tokens
+	"semantic_similarity_s1_v1":    0.0715, // ¥0.001 / 1k tokens
+	"hunyuan":                      7.143,  // ¥0.1 / 1k tokens  // https://cloud.tencent.com/document/product/1729/97731#e0e6be58-60c8-469f-bdeb-6c264ce3b4d0
+	// https://platform.lingyiwanwu.com/docs#-计费单元
+	// 已经按照 7.2 来换算美元价格
+	"yi-34b-chat-0205":      0.018,
+	"yi-34b-chat-200k":      0.0864,
+	"yi-vl-plus":            0.0432,
+	"command":               0.5,
+	"command-nightly":       0.5,
+	"command-light":         0.5,
+	"command-light-nightly": 0.5,
+	"command-r":             0.25,
+	"command-r-plus	":       1.5,
+	"deepseek-chat":         0.07,
+	"deepseek-coder":        0.07,
 }
 
 var DefaultModelPrice = map[string]float64{
@@ -114,14 +135,20 @@ var DefaultModelPrice = map[string]float64{
 	"swap_face":         0.05,
 }
 
-var ModelPrice = map[string]float64{}
-var ModelRatio = map[string]float64{}
+var modelPrice map[string]float64 = nil
+var modelRatio map[string]float64 = nil
+
+var CompletionRatio map[string]float64 = nil
+var DefaultCompletionRatio = map[string]float64{
+	"gpt-4-gizmo-*": 2,
+	"gpt-4-all":     2,
+}
 
 func ModelPrice2JSONString() string {
-	if len(ModelPrice) == 0 {
-		ModelPrice = DefaultModelPrice
+	if modelPrice == nil {
+		modelPrice = DefaultModelPrice
 	}
-	jsonBytes, err := json.Marshal(ModelPrice)
+	jsonBytes, err := json.Marshal(modelPrice)
 	if err != nil {
 		SysError("error marshalling model price: " + err.Error())
 	}
@@ -129,18 +156,18 @@ func ModelPrice2JSONString() string {
 }
 
 func UpdateModelPriceByJSONString(jsonStr string) error {
-	ModelPrice = make(map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &ModelPrice)
+	modelPrice = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &modelPrice)
 }
 
 func GetModelPrice(name string, printErr bool) float64 {
-	if len(ModelPrice) == 0 {
-		ModelPrice = DefaultModelPrice
+	if modelPrice == nil {
+		modelPrice = DefaultModelPrice
 	}
 	if strings.HasPrefix(name, "gpt-4-gizmo") {
 		name = "gpt-4-gizmo-*"
 	}
-	price, ok := ModelPrice[name]
+	price, ok := modelPrice[name]
 	if !ok {
 		if printErr {
 			SysError("model price not found: " + name)
@@ -151,10 +178,10 @@ func GetModelPrice(name string, printErr bool) float64 {
 }
 
 func ModelRatio2JSONString() string {
-	if len(ModelRatio) == 0 {
-		ModelRatio = DefaultModelRatio
+	if modelRatio == nil {
+		modelRatio = DefaultModelRatio
 	}
-	jsonBytes, err := json.Marshal(ModelRatio)
+	jsonBytes, err := json.Marshal(modelRatio)
 	if err != nil {
 		SysError("error marshalling model ratio: " + err.Error())
 	}
@@ -162,18 +189,18 @@ func ModelRatio2JSONString() string {
 }
 
 func UpdateModelRatioByJSONString(jsonStr string) error {
-	ModelRatio = make(map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &ModelRatio)
+	modelRatio = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &modelRatio)
 }
 
 func GetModelRatio(name string) float64 {
-	if len(ModelRatio) == 0 {
-		ModelRatio = DefaultModelRatio
+	if modelRatio == nil {
+		modelRatio = DefaultModelRatio
 	}
 	if strings.HasPrefix(name, "gpt-4-gizmo") {
 		name = "gpt-4-gizmo-*"
 	}
-	ratio, ok := ModelRatio[name]
+	ratio, ok := modelRatio[name]
 	if !ok {
 		SysError("model ratio not found: " + name)
 		return 30
@@ -181,37 +208,76 @@ func GetModelRatio(name string) float64 {
 	return ratio
 }
 
+func CompletionRatio2JSONString() string {
+	if CompletionRatio == nil {
+		CompletionRatio = DefaultCompletionRatio
+	}
+	jsonBytes, err := json.Marshal(CompletionRatio)
+	if err != nil {
+		SysError("error marshalling completion ratio: " + err.Error())
+	}
+	return string(jsonBytes)
+}
+
+func UpdateCompletionRatioByJSONString(jsonStr string) error {
+	CompletionRatio = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &CompletionRatio)
+}
+
 func GetCompletionRatio(name string) float64 {
 	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
 		}
 		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") {
+	if strings.HasPrefix(name, "gpt-4") && name != "gpt-4-all" && !strings.HasPrefix(name, "gpt-4-gizmo") {
+		if strings.HasPrefix(name, "gpt-4-turbo") || strings.HasSuffix(name, "preview") {
 			return 3
 		}
 		return 2
 	}
-	if strings.HasPrefix(name, "claude-instant-1") {
+	if strings.Contains(name, "claude-instant-1") {
 		return 3
-	} else if strings.HasPrefix(name, "claude-2") {
+	} else if strings.Contains(name, "claude-2") {
 		return 3
-	} else if strings.HasPrefix(name, "claude-3") {
+	} else if strings.Contains(name, "claude-3") {
 		return 5
 	}
+	if strings.HasPrefix(name, "mistral-") {
+		return 3
+	}
+	if strings.HasPrefix(name, "gemini-") {
+		return 3
+	}
+	if strings.HasPrefix(name, "command") {
+		switch name {
+		case "command-r":
+			return 3
+		case "command-r-plus":
+			return 5
+		default:
+			return 2
+		}
+	}
+	if strings.HasPrefix(name, "deepseek") {
+		return 2
+	}
+	switch name {
+	case "llama2-70b-4096":
+		return 0.8 / 0.64
+	case "llama3-8b-8192":
+		return 2
+	case "llama3-70b-8192":
+		return 0.79 / 0.59
+	}
+	if ratio, ok := CompletionRatio[name]; ok {
+		return ratio
+	}
 	return 1
 }

common/redis.go

@@ -18,9 +18,8 @@ func InitRedisClient() (err error) {
 		return nil
 	}
 	if os.Getenv("SYNC_FREQUENCY") == "" {
-		RedisEnabled = false
-		SysLog("SYNC_FREQUENCY not set, Redis is disabled")
-		return nil
+		SysLog("SYNC_FREQUENCY not set, use default value 60")
+		SyncFrequency = 60
 	}
 	SysLog("Redis is enabled")
 	opt, err := redis.ParseURL(os.Getenv("REDIS_CONN_STRING"))

common/utils.go

@@ -1,6 +1,7 @@
 package common
 
 import (
+	"encoding/json"
 	"fmt"
 	"github.com/google/uuid"
 	"html/template"
@@ -236,3 +237,16 @@ func StringToByteSlice(s string) []byte {
 	tmp2 := [3]uintptr{tmp1[0], tmp1[1], tmp1[1]}
 	return *(*[]byte)(unsafe.Pointer(&tmp2))
 }
+
+func RandomSleep() {
+	// Sleep for 0-3000 ms
+	time.Sleep(time.Duration(rand.Intn(3000)) * time.Millisecond)
+}
+
+func MapToJsonStr(m map[string]interface{}) string {
+	bytes, err := json.Marshal(m)
+	if err != nil {
+		return ""
+	}
+	return string(bytes)
+}

constant/midjourney.go

@@ -1,6 +1,9 @@
 package constant
 
 var MjNotifyEnabled = false
+var MjAccountFilterEnabled = false
+var MjModeClearEnabled = false
+var MjForwardUrlEnabled = true
 
 const (
 	MjErrorUnknown = 5

constant/payment.go

@@ -0,0 +1,8 @@
+package constant
+
+var PayAddress = ""
+var CustomCallbackAddress = ""
+var EpayId = ""
+var EpayKey = ""
+var Price = 7.3
+var MinTopUp = 1

constant/sensitive.go

@@ -4,7 +4,8 @@ import "strings"
 
 var CheckSensitiveEnabled = true
 var CheckSensitiveOnPromptEnabled = true
-var CheckSensitiveOnCompletionEnabled = true
+
+//var CheckSensitiveOnCompletionEnabled = true
 
 // StopOnSensitiveEnabled 如果检测到敏感词，是否立刻停止生成，否则替换敏感词
 var StopOnSensitiveEnabled = true
@@ -23,6 +24,7 @@ func SensitiveWordsToString() string {
 }
 
 func SensitiveWordsFromString(s string) {
+	SensitiveWords = []string{}
 	sw := strings.Split(s, "\n")
 	for _, w := range sw {
 		w = strings.TrimSpace(w)
@@ -36,6 +38,6 @@ func ShouldCheckPromptSensitive() bool {
 	return CheckSensitiveEnabled && CheckSensitiveOnPromptEnabled
 }
 
-func ShouldCheckCompletionSensitive() bool {
-	return CheckSensitiveEnabled && CheckSensitiveOnCompletionEnabled
-}
+//func ShouldCheckCompletionSensitive() bool {
+//	return CheckSensitiveEnabled && CheckSensitiveOnCompletionEnabled
+//}

controller/channel-test.go

@@ -27,7 +27,6 @@ func testChannel(channel *model.Channel, testModel string) (err error, openaiErr
 	if channel.Type == common.ChannelTypeMidjourney {
 		return errors.New("midjourney channel test is not supported"), nil
 	}
-	common.SysLog(fmt.Sprintf("testing channel %d with model %s", channel.Id, testModel))
 	w := httptest.NewRecorder()
 	c, _ := gin.CreateTestContext(w)
 	c.Request = &http.Request{
@@ -60,12 +59,16 @@ func testChannel(channel *model.Channel, testModel string) (err error, openaiErr
 		return fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), nil
 	}
 	if testModel == "" {
-		testModel = adaptor.GetModelList()[0]
-		meta.UpstreamModelName = testModel
+		if channel.TestModel != nil && *channel.TestModel != "" {
+			testModel = *channel.TestModel
+		} else {
+			testModel = adaptor.GetModelList()[0]
+		}
 	}
 	request := buildTestRequest()
 	request.Model = testModel
 	meta.UpstreamModelName = testModel
+	common.SysLog(fmt.Sprintf("testing channel %d with model %s", channel.Id, testModel))
 
 	adaptor.Init(meta, *request)
 
@@ -83,11 +86,11 @@ func testChannel(channel *model.Channel, testModel string) (err error, openaiErr
 	if err != nil {
 		return err, nil
 	}
-	if resp.StatusCode != http.StatusOK {
+	if resp != nil && resp.StatusCode != http.StatusOK {
 		err := relaycommon.RelayErrorHandler(resp)
 		return fmt.Errorf("status code %d: %s", resp.StatusCode, err.Error.Message), &err.Error
 	}
-	usage, respErr, _ := adaptor.DoResponse(c, resp, meta)
+	usage, respErr := adaptor.DoResponse(c, resp, meta)
 	if respErr != nil {
 		return fmt.Errorf("%s", respErr.Error.Message), &respErr.Error
 	}
@@ -108,6 +111,7 @@ func buildTestRequest() *dto.GeneralOpenAIRequest {
 	testRequest := &dto.GeneralOpenAIRequest{
 		Model:     "", // this will be set later
 		MaxTokens: 1,
+		Stream:    false,
 	}
 	content, _ := json.Marshal("hi")
 	testMessage := dto.Message{
@@ -204,7 +208,7 @@ func testAllChannels(notify bool) error {
 			if isChannelEnabled && service.ShouldDisableChannel(openaiErr, -1) && ban {
 				service.DisableChannel(channel.Id, channel.Name, err.Error())
 			}
-			if !isChannelEnabled && service.ShouldEnableChannel(err, openaiErr) {
+			if !isChannelEnabled && service.ShouldEnableChannel(err, openaiErr, channel.Status) {
 				service.EnableChannel(channel.Id, channel.Name)
 			}
 			channel.UpdateResponseTime(milliseconds)

controller/midjourney.go

@@ -10,11 +10,11 @@ import (
 	"log"
 	"net/http"
 	"one-api/common"
+	"one-api/constant"
 	"one-api/dto"
 	"one-api/model"
 	"one-api/service"
 	"strconv"
-	"strings"
 	"time"
 )
 
@@ -86,7 +86,7 @@ func UpdateMidjourneyTaskBulk() {
 				continue
 			}
 			// 设置超时时间
-			timeout := time.Second * 5
+			timeout := time.Second * 15
 			ctx, cancel := context.WithTimeout(context.Background(), timeout)
 			// 使用带有超时的 context 创建新的请求
 			req = req.WithContext(ctx)
@@ -147,7 +147,7 @@ func UpdateMidjourneyTaskBulk() {
 					task.Buttons = string(buttonStr)
 				}
 
-				if task.Progress != "100%" && responseItem.FailReason != "" {
+				if (task.Progress != "100%" && responseItem.FailReason != "") || (task.Progress == "100%" && task.Status == "FAILURE") {
 					common.LogInfo(ctx, task.MjId+" 构建失败，"+task.FailReason)
 					task.Progress = "100%"
 					err = model.CacheUpdateUserQuota(task.UserId)
@@ -233,6 +233,12 @@ func GetAllMidjourney(c *gin.Context) {
 	if logs == nil {
 		logs = make([]*model.Midjourney, 0)
 	}
+	if constant.MjForwardUrlEnabled {
+		for i, midjourney := range logs {
+			midjourney.ImageUrl = common.ServerAddress + "/mj/image/" + midjourney.MjId
+			logs[i] = midjourney
+		}
+	}
 	c.JSON(200, gin.H{
 		"success": true,
 		"message": "",
@@ -259,7 +265,7 @@ func GetUserMidjourney(c *gin.Context) {
 	if logs == nil {
 		logs = make([]*model.Midjourney, 0)
 	}
-	if !strings.Contains(common.ServerAddress, "localhost") {
+	if constant.MjForwardUrlEnabled {
 		for i, midjourney := range logs {
 			midjourney.ImageUrl = common.ServerAddress + "/mj/image/" + midjourney.MjId
 			logs[i] = midjourney

controller/misc.go

@@ -33,6 +33,7 @@ func GetStatus(c *gin.Context) {
 		"success": true,
 		"message": "",
 		"data": gin.H{
+			"version":                  common.Version,
 			"start_time":               common.StartTime,
 			"email_verification":       common.EmailVerificationEnabled,
 			"github_oauth":             common.GitHubOAuthEnabled,
@@ -45,8 +46,8 @@ func GetStatus(c *gin.Context) {
 			"wechat_qrcode":            common.WeChatAccountQRCodeImageURL,
 			"wechat_login":             common.WeChatAuthEnabled,
 			"server_address":           common.ServerAddress,
-			"price":                    common.Price,
-			"min_topup":                common.MinTopUp,
+			"price":                    constant.Price,
+			"min_topup":                constant.MinTopUp,
 			"turnstile_check":          common.TurnstileCheckEnabled,
 			"turnstile_site_key":       common.TurnstileSiteKey,
 			"top_up_link":              common.TopUpLink,
@@ -59,7 +60,7 @@ func GetStatus(c *gin.Context) {
 			"enable_data_export":       common.DataExportEnabled,
 			"data_export_default_time": common.DataExportDefaultTime,
 			"default_collapse_sidebar": common.DefaultCollapseSidebar,
-			"enable_online_topup":      common.PayAddress != "" && common.EpayId != "" && common.EpayKey != "",
+			"enable_online_topup":      constant.PayAddress != "" && constant.EpayId != "" && constant.EpayKey != "",
 			"mj_notify_enabled":        constant.MjNotifyEnabled,
 		},
 	})
@@ -119,10 +120,20 @@ func SendEmailVerification(c *gin.Context) {
 		})
 		return
 	}
+	parts := strings.Split(email, "@")
+	if len(parts) != 2 {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "无效的邮箱地址",
+		})
+		return
+	}
+	localPart := parts[0]
+	domainPart := parts[1]
 	if common.EmailDomainRestrictionEnabled {
 		allowed := false
 		for _, domain := range common.EmailDomainWhitelist {
-			if strings.HasSuffix(email, "@"+domain) {
+			if domainPart == domain {
 				allowed = true
 				break
 			}
@@ -130,11 +141,22 @@ func SendEmailVerification(c *gin.Context) {
 		if !allowed {
 			c.JSON(http.StatusOK, gin.H{
 				"success": false,
-				"message": "管理员启用了邮箱域名白名单，您的邮箱地址的域名不在白名单中",
+				"message": "The administrator has enabled the email domain name whitelist, and your email address is not allowed due to special symbols or it's not in the whitelist.",
 			})
 			return
 		}
 	}
+	if common.EmailAliasRestrictionEnabled {
+		containsSpecialSymbols := strings.Contains(localPart, "+") || strings.Contains(localPart, ".")
+		if containsSpecialSymbols {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "管理员已启用邮箱地址别名限制，您的邮箱地址由于包含特殊符号而被拒绝。",
+			})
+			return
+		}
+	}
+
 	if model.IsEmailAlreadyTaken(email) {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,

controller/model.go

@@ -4,12 +4,15 @@ import (
 	"fmt"
 	"github.com/gin-gonic/gin"
 	"net/http"
+	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
 	"one-api/model"
 	"one-api/relay"
 	"one-api/relay/channel/ai360"
+	"one-api/relay/channel/lingyiwanwu"
 	"one-api/relay/channel/moonshot"
+	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
 )
 
@@ -42,8 +45,9 @@ type OpenAIModels struct {
 
 var openAIModels []OpenAIModels
 var openAIModelsMap map[string]OpenAIModels
+var channelId2Models map[int][]string
 
-func init() {
+func getPermission() []OpenAIModelPermission {
 	var permission []OpenAIModelPermission
 	permission = append(permission, OpenAIModelPermission{
 		Id:                 "modelperm-LwHkVFn8AcMItP432fKKDIKJ",
@@ -59,7 +63,12 @@ func init() {
 		Group:              nil,
 		IsBlocking:         false,
 	})
+	return permission
+}
+
+func init() {
 	// https://platform.openai.com/docs/models/model-endpoint-compatibility
+	permission := getPermission()
 	for i := 0; i < relayconstant.APITypeDummy; i++ {
 		if i == relayconstant.APITypeAIProxyLibrary {
 			continue
@@ -84,7 +93,7 @@ func init() {
 			Id:         modelName,
 			Object:     "model",
 			Created:    1626777600,
-			OwnedBy:    "360",
+			OwnedBy:    ai360.ChannelName,
 			Permission: permission,
 			Root:       modelName,
 			Parent:     nil,
@@ -101,6 +110,17 @@ func init() {
 			Parent:     nil,
 		})
 	}
+	for _, modelName := range lingyiwanwu.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
+			Object:     "model",
+			Created:    1626777600,
+			OwnedBy:    "lingyiwanwu",
+			Permission: permission,
+			Root:       modelName,
+			Parent:     nil,
+		})
+	}
 	for modelName, _ := range constant.MidjourneyModel2Action {
 		openAIModels = append(openAIModels, OpenAIModels{
 			Id:         modelName,
@@ -116,6 +136,17 @@ func init() {
 	for _, model := range openAIModels {
 		openAIModelsMap[model.Id] = model
 	}
+	channelId2Models = make(map[int][]string)
+	for i := 1; i <= common.ChannelTypeDummy; i++ {
+		apiType := relayconstant.ChannelType2APIType(i)
+		if apiType == -1 || apiType == relayconstant.APITypeAIProxyLibrary {
+			continue
+		}
+		meta := &relaycommon.RelayInfo{ChannelType: i}
+		adaptor := relay.GetAdaptor(apiType)
+		adaptor.Init(meta, dto.GeneralOpenAIRequest{})
+		channelId2Models[i] = adaptor.GetModelList()
+	}
 }
 
 func ListModels(c *gin.Context) {
@@ -130,21 +161,39 @@ func ListModels(c *gin.Context) {
 	}
 	models := model.GetGroupModels(user.Group)
 	userOpenAiModels := make([]OpenAIModels, 0)
+	permission := getPermission()
 	for _, s := range models {
 		if _, ok := openAIModelsMap[s]; ok {
 			userOpenAiModels = append(userOpenAiModels, openAIModelsMap[s])
+		} else {
+			userOpenAiModels = append(userOpenAiModels, OpenAIModels{
+				Id:         s,
+				Object:     "model",
+				Created:    1626777600,
+				OwnedBy:    "openai",
+				Permission: permission,
+				Root:       s,
+				Parent:     nil,
+			})
 		}
 	}
 	c.JSON(200, gin.H{
-		"object": "list",
-		"data":   userOpenAiModels,
+		"success": true,
+		"data":    userOpenAiModels,
 	})
 }
 
 func ChannelListModels(c *gin.Context) {
 	c.JSON(200, gin.H{
-		"object": "list",
-		"data":   openAIModels,
+		"success": true,
+		"data":    openAIModels,
+	})
+}
+
+func DashboardListModels(c *gin.Context) {
+	c.JSON(200, gin.H{
+		"success": true,
+		"data":    channelId2Models,
 	})
 }
 

controller/option.go

@@ -14,7 +14,7 @@ func GetOptions(c *gin.Context) {
 	var options []*model.Option
 	common.OptionMapRWMutex.Lock()
 	for k, v := range common.OptionMap {
-		if strings.HasSuffix(k, "Token") || strings.HasSuffix(k, "Secret") {
+		if strings.HasSuffix(k, "Token") || strings.HasSuffix(k, "Secret") || strings.HasSuffix(k, "Key") {
 			continue
 		}
 		options = append(options, &model.Option{

controller/relay.go

@@ -1,21 +1,24 @@
 package controller
 
 import (
+	"bytes"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"io"
 	"log"
 	"net/http"
 	"one-api/common"
 	"one-api/dto"
+	"one-api/middleware"
+	"one-api/model"
 	"one-api/relay"
 	"one-api/relay/constant"
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
-	"strconv"
+	"strings"
 )
 
-func Relay(c *gin.Context) {
-	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
+func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 	var err *dto.OpenAIErrorWithStatusCode
 	switch relayMode {
 	case relayconstant.RelayModeImagesGenerations:
@@ -29,33 +32,102 @@ func Relay(c *gin.Context) {
 	default:
 		err = relay.TextHelper(c)
 	}
-	if err != nil {
-		requestId := c.GetString(common.RequestIdKey)
-		retryTimesStr := c.Query("retry")
-		retryTimes, _ := strconv.Atoi(retryTimesStr)
-		if retryTimesStr == "" {
-			retryTimes = common.RetryTimes
+	return err
+}
+
+func Relay(c *gin.Context) {
+	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
+	retryTimes := common.RetryTimes
+	requestId := c.GetString(common.RequestIdKey)
+	channelId := c.GetInt("channel_id")
+	group := c.GetString("group")
+	originalModel := c.GetString("original_model")
+	openaiErr := relayHandler(c, relayMode)
+	useChannel := []int{channelId}
+	if openaiErr != nil {
+		go processChannelError(c, channelId, openaiErr)
+	} else {
+		retryTimes = 0
+	}
+	for i := 0; shouldRetry(c, channelId, openaiErr, retryTimes) && i < retryTimes; i++ {
+		channel, err := model.CacheGetRandomSatisfiedChannel(group, originalModel, i)
+		if err != nil {
+			common.LogError(c.Request.Context(), fmt.Sprintf("CacheGetRandomSatisfiedChannel failed: %s", err.Error()))
+			break
 		}
-		if retryTimes > 0 {
-			c.Redirect(http.StatusTemporaryRedirect, fmt.Sprintf("%s?retry=%d", c.Request.URL.Path, retryTimes-1))
-		} else {
-			if err.StatusCode == http.StatusTooManyRequests {
-				//err.Error.Message = "当前分组上游负载已饱和，请稍后再试"
-			}
-			err.Error.Message = common.MessageWithRequestId(err.Error.Message, requestId)
-			c.JSON(err.StatusCode, gin.H{
-				"error": err.Error,
-			})
+		channelId = channel.Id
+		useChannel = append(useChannel, channelId)
+		common.LogInfo(c.Request.Context(), fmt.Sprintf("using channel #%d to retry (remain times %d)", channel.Id, i))
+		middleware.SetupContextForSelectedChannel(c, channel, originalModel)
+
+		requestBody, err := common.GetRequestBody(c)
+		c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
+		openaiErr = relayHandler(c, relayMode)
+		if openaiErr != nil {
+			go processChannelError(c, channelId, openaiErr)
 		}
-		channelId := c.GetInt("channel_id")
-		autoBan := c.GetBool("auto_ban")
-		common.LogError(c.Request.Context(), fmt.Sprintf("relay error (channel #%d): %s", channelId, err.Error.Message))
-		// https://platform.openai.com/docs/guides/error-codes/api-errors
-		if service.ShouldDisableChannel(&err.Error, err.StatusCode) && autoBan {
-			channelId := c.GetInt("channel_id")
-			channelName := c.GetString("channel_name")
-			service.DisableChannel(channelId, channelName, err.Error.Message)
+	}
+	if len(useChannel) > 1 {
+		retryLogStr := fmt.Sprintf("重试：%s", strings.Trim(strings.Join(strings.Fields(fmt.Sprint(useChannel)), "->"), "[]"))
+		common.LogInfo(c.Request.Context(), retryLogStr)
+	}
+
+	if openaiErr != nil {
+		if openaiErr.StatusCode == http.StatusTooManyRequests {
+			openaiErr.Error.Message = "当前分组上游负载已饱和，请稍后再试"
 		}
+		openaiErr.Error.Message = common.MessageWithRequestId(openaiErr.Error.Message, requestId)
+		c.JSON(openaiErr.StatusCode, gin.H{
+			"error": openaiErr.Error,
+		})
+	}
+}
+
+func shouldRetry(c *gin.Context, channelId int, openaiErr *dto.OpenAIErrorWithStatusCode, retryTimes int) bool {
+	if openaiErr == nil {
+		return false
+	}
+	if retryTimes <= 0 {
+		return false
+	}
+	if _, ok := c.Get("specific_channel_id"); ok {
+		return false
+	}
+	if openaiErr.StatusCode == http.StatusTooManyRequests {
+		return true
+	}
+	if openaiErr.StatusCode == 307 {
+		return true
+	}
+	if openaiErr.StatusCode/100 == 5 {
+		// 超时不重试
+		if openaiErr.StatusCode == 504 || openaiErr.StatusCode == 524 {
+			return false
+		}
+		return true
+	}
+	if openaiErr.StatusCode == http.StatusBadRequest {
+		return false
+	}
+	if openaiErr.StatusCode == 408 {
+		// azure处理超时不重试
+		return false
+	}
+	if openaiErr.LocalError {
+		return false
+	}
+	if openaiErr.StatusCode/100 == 2 {
+		return false
+	}
+	return true
+}
+
+func processChannelError(c *gin.Context, channelId int, err *dto.OpenAIErrorWithStatusCode) {
+	autoBan := c.GetBool("auto_ban")
+	common.LogError(c.Request.Context(), fmt.Sprintf("relay error (channel #%d, status code: %d): %s", channelId, err.StatusCode, err.Error.Message))
+	if service.ShouldDisableChannel(&err.Error, err.StatusCode) && autoBan {
+		channelName := c.GetString("channel_name")
+		service.DisableChannel(channelId, channelName, err.Error.Message)
 	}
 }
 
@@ -88,7 +160,7 @@ func RelayMidjourney(c *gin.Context) {
 			"code":        err.Code,
 		})
 		channelId := c.GetInt("channel_id")
-		common.SysError(fmt.Sprintf("relay error (channel #%d): %s", channelId, fmt.Sprintf("%s %s", err.Description, err.Result)))
+		common.LogError(c, fmt.Sprintf("relay error (channel #%d, status code %d): %s", channelId, statusCode, fmt.Sprintf("%s %s", err.Description, err.Result)))
 	}
 }
 

controller/topup.go

@@ -2,9 +2,11 @@ package controller
 
 import (
 	"fmt"
+	"github.com/Calcium-Ion/go-epay/epay"
 	"github.com/gin-gonic/gin"
 	"github.com/samber/lo"
-	epay "github.com/star-horizon/go-epay"
+	"one-api/constant"
+
 	"log"
 	"net/url"
 	"one-api/common"
@@ -27,44 +29,59 @@ type AmountRequest struct {
 }
 
 func GetEpayClient() *epay.Client {
-	if common.PayAddress == "" || common.EpayId == "" || common.EpayKey == "" {
+	if constant.PayAddress == "" || constant.EpayId == "" || constant.EpayKey == "" {
 		return nil
 	}
-	withUrl, err := epay.NewClientWithUrl(&epay.Config{
-		PartnerID: common.EpayId,
-		Key:       common.EpayKey,
-	}, common.PayAddress)
+	withUrl, err := epay.NewClient(&epay.Config{
+		PartnerID: constant.EpayId,
+		Key:       constant.EpayKey,
+	}, constant.PayAddress)
 	if err != nil {
 		return nil
 	}
 	return withUrl
 }
 
-func GetAmount(count float64, user model.User) float64 {
+func getPayMoney(amount float64, user model.User) float64 {
+	if !common.DisplayInCurrencyEnabled {
+		amount = amount / common.QuotaPerUnit
+	}
 	// 别问为什么用float64，问就是这么点钱没必要
 	topupGroupRatio := common.GetTopupGroupRatio(user.Group)
 	if topupGroupRatio == 0 {
 		topupGroupRatio = 1
 	}
-	amount := count * common.Price * topupGroupRatio
-	return amount
+	payMoney := amount * constant.Price * topupGroupRatio
+	return payMoney
+}
+
+func getMinTopup() int {
+	minTopup := constant.MinTopUp
+	if !common.DisplayInCurrencyEnabled {
+		minTopup = minTopup * int(common.QuotaPerUnit)
+	}
+	return minTopup
 }
 
 func RequestEpay(c *gin.Context) {
 	var req EpayRequest
 	err := c.ShouldBindJSON(&req)
 	if err != nil {
-		c.JSON(200, gin.H{"message": err.Error(), "data": 10})
+		c.JSON(200, gin.H{"message": "error", "data": "参数错误"})
 		return
 	}
-	if req.Amount < common.MinTopUp {
-		c.JSON(200, gin.H{"message": fmt.Sprintf("充值数量不能小于 %d", common.MinTopUp), "data": 10})
+	if req.Amount < getMinTopup() {
+		c.JSON(200, gin.H{"message": "error", "data": fmt.Sprintf("充值数量不能小于 %d", getMinTopup())})
 		return
 	}
 
 	id := c.GetInt("id")
 	user, _ := model.GetUserById(id, false)
-	payMoney := GetAmount(float64(req.Amount), *user)
+	payMoney := getPayMoney(float64(req.Amount), *user)
+	if payMoney < 0.01 {
+		c.JSON(200, gin.H{"message": "error", "data": "充值金额过低"})
+		return
+	}
 
 	var payType epay.PurchaseType
 	if req.PaymentMethod == "zfb" {
@@ -77,7 +94,7 @@ func RequestEpay(c *gin.Context) {
 	callBackAddress := service.GetCallbackAddress()
 	returnUrl, _ := url.Parse(common.ServerAddress + "/log")
 	notifyUrl, _ := url.Parse(callBackAddress + "/api/user/epay/notify")
-	tradeNo := strconv.FormatInt(time.Now().Unix(), 10)
+	tradeNo := fmt.Sprintf("%s%d", common.GetRandomString(6), time.Now().Unix())
 	client := GetEpayClient()
 	if client == nil {
 		c.JSON(200, gin.H{"message": "error", "data": "当前管理员未配置支付信息"})
@@ -96,9 +113,13 @@ func RequestEpay(c *gin.Context) {
 		c.JSON(200, gin.H{"message": "error", "data": "拉起支付失败"})
 		return
 	}
+	amount := req.Amount
+	if !common.DisplayInCurrencyEnabled {
+		amount = amount / int(common.QuotaPerUnit)
+	}
 	topUp := &model.TopUp{
 		UserId:     id,
-		Amount:     req.Amount,
+		Amount:     amount,
 		Money:      payMoney,
 		TradeNo:    "A" + tradeNo,
 		CreateTime: time.Now().Unix(),
@@ -186,13 +207,13 @@ func EpayNotify(c *gin.Context) {
 			}
 			//user, _ := model.GetUserById(topUp.UserId, false)
 			//user.Quota += topUp.Amount * 500000
-			err = model.IncreaseUserQuota(topUp.UserId, topUp.Amount*500000)
+			err = model.IncreaseUserQuota(topUp.UserId, topUp.Amount*int(common.QuotaPerUnit))
 			if err != nil {
 				log.Printf("易支付回调更新用户失败: %v", topUp)
 				return
 			}
 			log.Printf("易支付回调更新用户成功 %v", topUp)
-			model.RecordLog(topUp.UserId, model.LogTypeTopup, fmt.Sprintf("使用在线充值成功，充值金额: %v，支付金额：%f", common.LogQuota(topUp.Amount*500000), topUp.Money))
+			model.RecordLog(topUp.UserId, model.LogTypeTopup, fmt.Sprintf("使用在线充值成功，充值金额: %v，支付金额：%f", common.LogQuota(topUp.Amount*int(common.QuotaPerUnit)), topUp.Money))
 		}
 	} else {
 		log.Printf("易支付异常回调: %v", verifyInfo)
@@ -206,12 +227,17 @@ func RequestAmount(c *gin.Context) {
 		c.JSON(200, gin.H{"message": "error", "data": "参数错误"})
 		return
 	}
-	if req.Amount < common.MinTopUp {
-		c.JSON(200, gin.H{"message": "error", "data": fmt.Sprintf("充值数量不能小于 %d", common.MinTopUp)})
+
+	if req.Amount < getMinTopup() {
+		c.JSON(200, gin.H{"message": "error", "data": fmt.Sprintf("充值数量不能小于 %d", getMinTopup())})
 		return
 	}
 	id := c.GetInt("id")
 	user, _ := model.GetUserById(id, false)
-	payMoney := GetAmount(float64(req.Amount), *user)
+	payMoney := getPayMoney(float64(req.Amount), *user)
+	if payMoney <= 0.01 {
+		c.JSON(200, gin.H{"message": "error", "data": "充值金额过低"})
+		return
+	}
 	c.JSON(200, gin.H{"message": "success", "data": strconv.FormatFloat(payMoney, 'f', 2, 64)})
 }

controller/user.go

@@ -7,6 +7,7 @@ import (
 	"one-api/common"
 	"one-api/model"
 	"strconv"
+	"sync"
 
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
@@ -215,7 +216,8 @@ func GetAllUsers(c *gin.Context) {
 
 func SearchUsers(c *gin.Context) {
 	keyword := c.Query("keyword")
-	users, err := model.SearchUsers(keyword)
+	group := c.Query("group")
+	users, err := model.SearchUsers(keyword, group)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -451,7 +453,7 @@ func UpdateUser(c *gin.Context) {
 		updatedUser.Password = "" // rollback to what it should be
 	}
 	updatePassword := updatedUser.Password != ""
-	if err := updatedUser.Update(updatePassword); err != nil {
+	if err := updatedUser.Edit(updatePassword); err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
 			"message": err.Error(),
@@ -789,7 +791,11 @@ type topUpRequest struct {
 	Key string `json:"key"`
 }
 
+var lock = sync.Mutex{}
+
 func TopUp(c *gin.Context) {
+	lock.Lock()
+	defer lock.Unlock()
 	req := topUpRequest{}
 	err := c.ShouldBindJSON(&req)
 	if err != nil {

dto/error.go

@@ -10,6 +10,7 @@ type OpenAIError struct {
 type OpenAIErrorWithStatusCode struct {
 	Error      OpenAIError `json:"error"`
 	StatusCode int         `json:"status_code"`
+	LocalError bool
 }
 
 type GeneralErrorResponse struct {

dto/text_request.go

@@ -32,6 +32,21 @@ type GeneralOpenAIRequest struct {
 	TopLogProbs      int             `json:"top_logprobs,omitempty"`
 }
 
+type OpenAITools struct {
+	Type     string         `json:"type"`
+	Function OpenAIFunction `json:"function"`
+}
+
+type OpenAIFunction struct {
+	Description string `json:"description,omitempty"`
+	Name        string `json:"name"`
+	Parameters  any    `json:"parameters,omitempty"`
+}
+
+func (r GeneralOpenAIRequest) GetMaxTokens() int64 {
+	return int64(r.MaxTokens)
+}
+
 func (r GeneralOpenAIRequest) ParseInput() []string {
 	if r.Input == nil {
 		return nil

dto/text_response.go

@@ -1,9 +1,29 @@
 package dto
 
-type TextResponse struct {
-	Choices []*OpenAITextResponseChoice `json:"choices"`
+type TextResponseWithError struct {
+	Id      string                        `json:"id"`
+	Object  string                        `json:"object"`
+	Created int64                         `json:"created"`
+	Choices []OpenAITextResponseChoice    `json:"choices"`
+	Data    []OpenAIEmbeddingResponseItem `json:"data"`
+	Model   string                        `json:"model"`
+	Usage   `json:"usage"`
+	Error   OpenAIError `json:"error"`
+}
+
+type SimpleResponse struct {
+	Usage   `json:"usage"`
+	Error   OpenAIError                `json:"error"`
+	Choices []OpenAITextResponseChoice `json:"choices"`
+}
+
+type TextResponse struct {
+	Id      string                     `json:"id"`
+	Object  string                     `json:"object"`
+	Created int64                      `json:"created"`
+	Model   string                     `json:"model"`
+	Choices []OpenAITextResponseChoice `json:"choices"`
 	Usage   `json:"usage"`
-	Error   *OpenAIError `json:"error,omitempty"`
 }
 
 type OpenAITextResponseChoice struct {
@@ -34,21 +54,54 @@ type OpenAIEmbeddingResponse struct {
 }
 
 type ChatCompletionsStreamResponseChoice struct {
-	Delta struct {
-		Content   string `json:"content"`
-		Role      string `json:"role,omitempty"`
-		ToolCalls any    `json:"tool_calls,omitempty"`
-	} `json:"delta"`
-	FinishReason *string `json:"finish_reason,omitempty"`
-	Index        int     `json:"index,omitempty"`
+	Delta        ChatCompletionsStreamResponseChoiceDelta `json:"delta,omitempty"`
+	Logprobs     *any                                     `json:"logprobs"`
+	FinishReason *string                                  `json:"finish_reason"`
+	Index        int                                      `json:"index"`
+}
+
+type ChatCompletionsStreamResponseChoiceDelta struct {
+	Content   *string    `json:"content,omitempty"`
+	Role      string     `json:"role,omitempty"`
+	ToolCalls []ToolCall `json:"tool_calls,omitempty"`
+}
+
+func (c *ChatCompletionsStreamResponseChoiceDelta) IsEmpty() bool {
+	return c.Content == nil && len(c.ToolCalls) == 0
+}
+
+func (c *ChatCompletionsStreamResponseChoiceDelta) SetContentString(s string) {
+	c.Content = &s
+}
+
+func (c *ChatCompletionsStreamResponseChoiceDelta) GetContentString() string {
+	if c.Content == nil {
+		return ""
+	}
+	return *c.Content
+}
+
+type ToolCall struct {
+	// Index is not nil only in chat completion chunk object
+	Index    *int         `json:"index,omitempty"`
+	ID       string       `json:"id"`
+	Type     any          `json:"type"`
+	Function FunctionCall `json:"function"`
+}
+
+type FunctionCall struct {
+	Name string `json:"name,omitempty"`
+	// call function with arguments in JSON format
+	Arguments string `json:"arguments,omitempty"`
 }
 
 type ChatCompletionsStreamResponse struct {
-	Id      string                                `json:"id"`
-	Object  string                                `json:"object"`
-	Created int64                                 `json:"created"`
-	Model   string                                `json:"model"`
-	Choices []ChatCompletionsStreamResponseChoice `json:"choices"`
+	Id                string                                `json:"id"`
+	Object            string                                `json:"object"`
+	Created           int64                                 `json:"created"`
+	Model             string                                `json:"model"`
+	SystemFingerprint *string                               `json:"system_fingerprint"`
+	Choices           []ChatCompletionsStreamResponseChoice `json:"choices"`
 }
 
 type ChatCompletionsStreamResponseSimple struct {

go.mod

@@ -4,7 +4,11 @@ module one-api
 go 1.18
 
 require (
+	github.com/Calcium-Ion/go-epay v0.0.2
 	github.com/anknown/ahocorasick v0.0.0-20190904063843-d75dbd5169c0
+	github.com/aws/aws-sdk-go-v2 v1.26.1
+	github.com/aws/aws-sdk-go-v2/credentials v1.17.11
+	github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4
 	github.com/gin-contrib/cors v1.4.0
 	github.com/gin-contrib/gzip v0.0.6
 	github.com/gin-contrib/sessions v0.0.5
@@ -15,10 +19,11 @@ require (
 	github.com/golang-jwt/jwt v3.2.2+incompatible
 	github.com/google/uuid v1.3.0
 	github.com/gorilla/websocket v1.5.0
+	github.com/jinzhu/copier v0.4.0
+	github.com/pkg/errors v0.9.1
 	github.com/pkoukk/tiktoken-go v0.1.6
-	github.com/samber/lo v1.38.1
+	github.com/samber/lo v1.39.0
 	github.com/shirou/gopsutil v3.21.11+incompatible
-	github.com/star-horizon/go-epay v0.0.0-20230204124159-fa2e2293fdc2
 	golang.org/x/crypto v0.21.0
 	golang.org/x/image v0.15.0
 	gorm.io/driver/mysql v1.4.3
@@ -29,6 +34,10 @@ require (
 
 require (
 	github.com/anknown/darts v0.0.0-20151216065714-83ff685239e6 // indirect
+	github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5 // indirect
+	github.com/aws/smithy-go v1.20.2 // indirect
 	github.com/bytedance/sonic v1.9.1 // indirect
 	github.com/cespare/xxhash/v2 v2.1.2 // indirect
 	github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 // indirect
@@ -65,9 +74,9 @@ require (
 	github.com/ugorji/go/codec v1.2.11 // indirect
 	github.com/yusufpapurcu/wmi v1.2.3 // indirect
 	golang.org/x/arch v0.3.0 // indirect
-	golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17 // indirect
+	golang.org/x/exp v0.0.0-20240404231335-c0f41cb1a7a0 // indirect
 	golang.org/x/net v0.21.0 // indirect
-	golang.org/x/sync v0.1.0 // indirect
+	golang.org/x/sync v0.7.0 // indirect
 	golang.org/x/sys v0.18.0 // indirect
 	golang.org/x/text v0.14.0 // indirect
 	google.golang.org/protobuf v1.30.0 // indirect

go.sum

@@ -1,7 +1,23 @@
+github.com/Calcium-Ion/go-epay v0.0.2 h1:3knFBuaBFpHzsGeGQU/QxUqZSHh5s0+jGo0P62pJzWc=
+github.com/Calcium-Ion/go-epay v0.0.2/go.mod h1:cxo/ZOg8ClvE3VAnCmEzbuyAZINSq7kFEN9oHj5WQ2U=
 github.com/anknown/ahocorasick v0.0.0-20190904063843-d75dbd5169c0 h1:onfun1RA+KcxaMk1lfrRnwCd1UUuOjJM/lri5eM1qMs=
 github.com/anknown/ahocorasick v0.0.0-20190904063843-d75dbd5169c0/go.mod h1:4yg+jNTYlDEzBjhGS96v+zjyA3lfXlFd5CiTLIkPBLI=
 github.com/anknown/darts v0.0.0-20151216065714-83ff685239e6 h1:HblK3eJHq54yET63qPCTJnks3loDse5xRmmqHgHzwoI=
 github.com/anknown/darts v0.0.0-20151216065714-83ff685239e6/go.mod h1:pbiaLIeYLUbgMY1kwEAdwO6UKD5ZNwdPGQlwokS9fe8=
+github.com/aws/aws-sdk-go-v2 v1.26.1 h1:5554eUqIYVWpU0YmeeYZ0wU64H2VLBs8TlhRB2L+EkA=
+github.com/aws/aws-sdk-go-v2 v1.26.1/go.mod h1:ffIFB97e2yNsv4aTSGkqtHnppsIJzw7G7BReUZ3jCXM=
+github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2 h1:x6xsQXGSmW6frevwDA+vi/wqhp1ct18mVXYN08/93to=
+github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.6.2/go.mod h1:lPprDr1e6cJdyYeGXnRaJoP4Md+cDBvi2eOj00BlGmg=
+github.com/aws/aws-sdk-go-v2/credentials v1.17.11 h1:YuIB1dJNf1Re822rriUOTxopaHHvIq0l/pX3fwO+Tzs=
+github.com/aws/aws-sdk-go-v2/credentials v1.17.11/go.mod h1:AQtFPsDH9bI2O+71anW6EKL+NcD7LG3dpKGMV4SShgo=
+github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5 h1:aw39xVGeRWlWx9EzGVnhOR4yOjQDHPQ6o6NmBlscyQg=
+github.com/aws/aws-sdk-go-v2/internal/configsources v1.3.5/go.mod h1:FSaRudD0dXiMPK2UjknVwwTYyZMRsHv3TtkabsZih5I=
+github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5 h1:PG1F3OD1szkuQPzDw3CIQsRIrtTlUC3lP84taWzHlq0=
+github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.6.5/go.mod h1:jU1li6RFryMz+so64PpKtudI+QzbKoIEivqdf6LNpOc=
+github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4 h1:JgHnonzbnA3pbqj76wYsSZIZZQYBxkmMEjvL6GHy8XU=
+github.com/aws/aws-sdk-go-v2/service/bedrockruntime v1.7.4/go.mod h1:nZspkhg+9p8iApLFoyAqfyuMP0F38acy2Hm3r5r95Cg=
+github.com/aws/smithy-go v1.20.2 h1:tbp628ireGtzcHDDmLT/6ADHidqnwgF57XOXZe6tp4Q=
+github.com/aws/smithy-go v1.20.2/go.mod h1:krry+ya/rV9RDcV/Q16kpu6ypI4K2czasz0NC3qS14E=
 github.com/bytedance/sonic v1.5.0/go.mod h1:ED5hyg4y6t3/9Ku1R6dU/4KyJ48DZ4jPhfY1O2AihPM=
 github.com/bytedance/sonic v1.9.1 h1:6iJ6NqdoxCDr6mbY8h18oSO+cShGSMRGCEo7F2h0x8s=
 github.com/bytedance/sonic v1.9.1/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZXU064P/U=
@@ -62,8 +78,8 @@ 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 v0.5.8 h1:e6P7q2lk1O+qJJb4BtCQXlK8vWEO8V1ZeuEdJNOqZyg=
 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=
@@ -83,6 +99,8 @@ github.com/jackc/pgx/v5 v5.5.1 h1:5I9etrGkLrN+2XPCsi6XLlV5DITbSL/xBZdmAxFcXPI=
 github.com/jackc/pgx/v5 v5.5.1/go.mod h1:Ig06C2Vu0t5qXC60W8sqIthScaEnFvojjj9dSljmHRA=
 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/copier v0.4.0 h1:w3ciUoD19shMCRargcpm0cm91ytaBhDvuRpz1ODO/U8=
+github.com/jinzhu/copier v0.4.0/go.mod h1:DfbEm0FYsaqBcKcFuvmOZb218JkPGtvSHsKg8S8hyyg=
 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=
@@ -128,6 +146,8 @@ github.com/pelletier/go-toml/v2 v2.0.1/go.mod h1:r9LEWfGN8R5k0VXJ+0BkIe7MYkRdwZO
 github.com/pelletier/go-toml/v2 v2.0.8 h1:0ctb6s9mE31h0/lhu+J6OPmVeDxJn+kYnJc2jZR9tGQ=
 github.com/pelletier/go-toml/v2 v2.0.8/go.mod h1:vuYfssBdrU2XDZ9bYydBu6t+6a6PYNcZljzZR9VXg+4=
 github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e/go.mod h1:pJLUxLENpZxwdsKMEsNbx1VGcRFpLqf3715MtcvvzbA=
+github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
+github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkoukk/tiktoken-go v0.1.6 h1:JF0TlJzhTbrI30wCvFuiw6FzP2+/bR+FIxUdgEAcUsw=
 github.com/pkoukk/tiktoken-go v0.1.6/go.mod h1:9NiV+i9mJKGj1rYOT+njbv+ZwA/zJxYdewGl6qVatpg=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
@@ -135,12 +155,10 @@ 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/samber/lo v1.38.1 h1:j2XEAqXKb09Am4ebOg31SpvzUTTs6EN3VfgeLUhPdXM=
-github.com/samber/lo v1.38.1/go.mod h1:+m/ZKRl6ClXCE2Lgf3MsQlWfh4bn1bz6CXEOxnEXnEA=
+github.com/samber/lo v1.39.0 h1:4gTz1wUhNYLhFSKl6O+8peW0v2F4BCY034GRpU9WnuA=
+github.com/samber/lo v1.39.0/go.mod h1:+m/ZKRl6ClXCE2Lgf3MsQlWfh4bn1bz6CXEOxnEXnEA=
 github.com/shirou/gopsutil v3.21.11+incompatible h1:+1+c1VGhc88SSonWP6foOcLhvnKlUeu/erjjvaPEYiI=
 github.com/shirou/gopsutil v3.21.11+incompatible/go.mod h1:5b4v6he4MtMOwMlS0TUMTu2PcXUg8+E1lC7eC3UO/RA=
-github.com/star-horizon/go-epay v0.0.0-20230204124159-fa2e2293fdc2 h1:avbt5a8F/zbYwFzTugrqWOBJe/K1cJj6+xpr+x1oVAI=
-github.com/star-horizon/go-epay v0.0.0-20230204124159-fa2e2293fdc2/go.mod h1:SiffGCWGGMVwujne2dUQbJ5zUVD1V1Yj0hDuTfqFNEo=
 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=
@@ -173,15 +191,15 @@ golang.org/x/arch v0.3.0/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/crypto v0.0.0-20210711020723-a769d52b0f97/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
 golang.org/x/crypto v0.21.0 h1:X31++rzVUdKhX5sWmSOFZxx8UW/ldWx55cbf08iNAMA=
 golang.org/x/crypto v0.21.0/go.mod h1:0BP7YvVV9gBbVKyeTG0Gyn+gZm94bibOW5BjDEYAOMs=
-golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17 h1:3MTrJm4PyNL9NBqvYDSj3DHl46qQakyfqfWo4jgfaEM=
-golang.org/x/exp v0.0.0-20220303212507-bbda1eaf7a17/go.mod h1:lgLbSvA5ygNOMpwM/9anMpWVlVJ7Z+cHWq/eFuinpGE=
+golang.org/x/exp v0.0.0-20240404231335-c0f41cb1a7a0 h1:985EYyeCOxTpcgOTJpflJUwOeEz0CQOdPt73OzpE9F8=
+golang.org/x/exp v0.0.0-20240404231335-c0f41cb1a7a0/go.mod h1:/lliqkxwWAhPjf5oSOIJup2XcqJaw8RGS6k3TGEc7GI=
 golang.org/x/image v0.15.0 h1:kOELfmgrmJlw4Cdb7g/QGuB3CvDrXbqEIww/pNtNBm8=
 golang.org/x/image v0.15.0/go.mod h1:HUYqC05R2ZcZ3ejNQsIHQDQiwWM4JBqmm6MKANTp4LE=
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=
 golang.org/x/net v0.21.0 h1:AQyQV4dYCvJ7vGmJyKki9+PBdyvhkSd8EIx/qb0AYv4=
 golang.org/x/net v0.21.0/go.mod h1:bIjVDfnllIU7BJ2DNgfnXvpSvtn8VRwhlsaeUTyUS44=
-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/sync v0.7.0 h1:YsImfSBoP9QPYL0xyKJPq0gcaJdG3rInoqxTWbfQu9M=
+golang.org/x/sync v0.7.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
 golang.org/x/sys v0.0.0-20190916202348-b4ddaad3f8a3/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 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=
@@ -202,7 +220,6 @@ 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/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=

main.go

@@ -20,10 +20,10 @@ import (
 	_ "net/http/pprof"
 )
 
-//go:embed web/build
+//go:embed web/dist
 var buildFS embed.FS
 
-//go:embed web/build/index.html
+//go:embed web/dist/index.html
 var indexPage []byte
 
 func main() {

makefile

@@ -7,7 +7,7 @@ all: build-frontend start-backend
 
 build-frontend:
 	@echo "Building frontend..."
-	@cd $(FRONTEND_DIR) && npm install && DISABLE_ESLINT_PLUGIN='true' REACT_APP_VERSION=$(cat VERSION) npm run build npm run build
+	@cd $(FRONTEND_DIR) && npm install && DISABLE_ESLINT_PLUGIN='true' VITE_REACT_APP_VERSION=$(cat VERSION) npm run build
 
 start-backend:
 	@echo "Starting backend dev server..."

middleware/auth.go

@@ -127,7 +127,7 @@ func TokenAuth() func(c *gin.Context) {
 		}
 		if len(parts) > 1 {
 			if model.IsAdmin(token.UserId) {
-				c.Set("channelId", parts[1])
+				c.Set("specific_channel_id", parts[1])
 			} else {
 				abortWithOpenAiMessage(c, http.StatusForbidden, "普通用户不支持指定渠道")
 				return

middleware/distributor.go

@@ -23,7 +23,10 @@ func Distribute() func(c *gin.Context) {
 	return func(c *gin.Context) {
 		userId := c.GetInt("id")
 		var channel *model.Channel
-		channelId, ok := c.Get("channelId")
+		channelId, ok := c.Get("specific_channel_id")
+		modelRequest, shouldSelectChannel, err := getModelRequest(c)
+		userGroup, _ := model.CacheGetUserGroup(userId)
+		c.Set("group", userGroup)
 		if ok {
 			id, err := strconv.Atoi(channelId.(string))
 			if err != nil {
@@ -40,72 +43,7 @@ func Distribute() func(c *gin.Context) {
 				return
 			}
 		} else {
-			shouldSelectChannel := true
 			// Select a channel for the user
-			var modelRequest ModelRequest
-			var err error
-			if strings.HasPrefix(c.Request.URL.Path, "/mj") {
-				relayMode := relayconstant.Path2RelayModeMidjourney(c.Request.URL.Path)
-				if relayMode == relayconstant.RelayModeMidjourneyTaskFetch ||
-					relayMode == relayconstant.RelayModeMidjourneyTaskFetchByCondition ||
-					relayMode == relayconstant.RelayModeMidjourneyNotify ||
-					relayMode == relayconstant.RelayModeMidjourneyTaskImageSeed {
-					shouldSelectChannel = false
-				} else {
-					midjourneyRequest := dto.MidjourneyRequest{}
-					err = common.UnmarshalBodyReusable(c, &midjourneyRequest)
-					if err != nil {
-						abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, "+err.Error())
-						return
-					}
-					midjourneyModel, mjErr, success := service.GetMjRequestModel(relayMode, &midjourneyRequest)
-					if mjErr != nil {
-						abortWithMidjourneyMessage(c, http.StatusBadRequest, mjErr.Code, mjErr.Description)
-						return
-					}
-					if midjourneyModel == "" {
-						if !success {
-							abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, 无法解析模型")
-							return
-						} else {
-							// task fetch, task fetch by condition, notify
-							shouldSelectChannel = false
-						}
-					}
-					modelRequest.Model = midjourneyModel
-				}
-				c.Set("relay_mode", relayMode)
-			} else if !strings.HasPrefix(c.Request.URL.Path, "/v1/audio/transcriptions") {
-				err = common.UnmarshalBodyReusable(c, &modelRequest)
-			}
-			if err != nil {
-				abortWithOpenAiMessage(c, http.StatusBadRequest, "无效的请求, "+err.Error())
-				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"
-				}
-			}
-			if strings.HasPrefix(c.Request.URL.Path, "/v1/audio") {
-				if modelRequest.Model == "" {
-					if strings.HasPrefix(c.Request.URL.Path, "/v1/audio/speech") {
-						modelRequest.Model = "tts-1"
-					} else {
-						modelRequest.Model = "whisper-1"
-					}
-				}
-			}
 			// check token model mapping
 			modelLimitEnable := c.GetBool("token_model_limit_enabled")
 			if modelLimitEnable {
@@ -128,10 +66,8 @@ func Distribute() func(c *gin.Context) {
 				}
 			}
 
-			userGroup, _ := model.CacheGetUserGroup(userId)
-			c.Set("group", userGroup)
 			if shouldSelectChannel {
-				channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, modelRequest.Model)
+				channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, modelRequest.Model, 0)
 				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)
 					// 如果错误，但是渠道不为空，说明是数据库一致性问题
@@ -147,36 +83,114 @@ func Distribute() func(c *gin.Context) {
 					abortWithOpenAiMessage(c, http.StatusServiceUnavailable, fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道（数据库一致性已被破坏）", userGroup, modelRequest.Model))
 					return
 				}
-				c.Set("channel", channel.Type)
-				c.Set("channel_id", channel.Id)
-				c.Set("channel_name", channel.Name)
-				ban := true
-				// parse *int to bool
-				if channel.AutoBan != nil && *channel.AutoBan == 0 {
-					ban = false
-				}
-				if nil != channel.OpenAIOrganization {
-					c.Set("channel_organization", *channel.OpenAIOrganization)
-				}
-				c.Set("auto_ban", ban)
-				c.Set("model_mapping", channel.GetModelMapping())
-				c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
-				c.Set("base_url", channel.GetBaseURL())
-				// TODO: api_version统一
-				switch channel.Type {
-				case common.ChannelTypeAzure:
-					c.Set("api_version", channel.Other)
-				case common.ChannelTypeXunfei:
-					c.Set("api_version", channel.Other)
-				//case common.ChannelTypeAIProxyLibrary:
-				//	c.Set("library_id", channel.Other)
-				case common.ChannelTypeGemini:
-					c.Set("api_version", channel.Other)
-				case common.ChannelTypeAli:
-					c.Set("plugin", channel.Other)
-				}
 			}
 		}
+		SetupContextForSelectedChannel(c, channel, modelRequest.Model)
 		c.Next()
 	}
 }
+
+func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
+	var modelRequest ModelRequest
+	shouldSelectChannel := true
+	var err error
+	if strings.Contains(c.Request.URL.Path, "/mj/") {
+		relayMode := relayconstant.Path2RelayModeMidjourney(c.Request.URL.Path)
+		if relayMode == relayconstant.RelayModeMidjourneyTaskFetch ||
+			relayMode == relayconstant.RelayModeMidjourneyTaskFetchByCondition ||
+			relayMode == relayconstant.RelayModeMidjourneyNotify ||
+			relayMode == relayconstant.RelayModeMidjourneyTaskImageSeed {
+			shouldSelectChannel = false
+		} else {
+			midjourneyRequest := dto.MidjourneyRequest{}
+			err = common.UnmarshalBodyReusable(c, &midjourneyRequest)
+			if err != nil {
+				abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, "+err.Error())
+				return nil, false, err
+			}
+			midjourneyModel, mjErr, success := service.GetMjRequestModel(relayMode, &midjourneyRequest)
+			if mjErr != nil {
+				abortWithMidjourneyMessage(c, http.StatusBadRequest, mjErr.Code, mjErr.Description)
+				return nil, false, fmt.Errorf(mjErr.Description)
+			}
+			if midjourneyModel == "" {
+				if !success {
+					abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, 无法解析模型")
+					return nil, false, fmt.Errorf("无效的请求, 无法解析模型")
+				} else {
+					// task fetch, task fetch by condition, notify
+					shouldSelectChannel = false
+				}
+			}
+			modelRequest.Model = midjourneyModel
+		}
+		c.Set("relay_mode", relayMode)
+	} else if !strings.HasPrefix(c.Request.URL.Path, "/v1/audio/transcriptions") {
+		err = common.UnmarshalBodyReusable(c, &modelRequest)
+	}
+	if err != nil {
+		abortWithOpenAiMessage(c, http.StatusBadRequest, "无效的请求, "+err.Error())
+		return nil, false, 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"
+		}
+	}
+	if strings.HasPrefix(c.Request.URL.Path, "/v1/audio") {
+		if modelRequest.Model == "" {
+			if strings.HasPrefix(c.Request.URL.Path, "/v1/audio/speech") {
+				modelRequest.Model = "tts-1"
+			} else {
+				modelRequest.Model = "whisper-1"
+			}
+		}
+	}
+	return &modelRequest, shouldSelectChannel, nil
+}
+
+func SetupContextForSelectedChannel(c *gin.Context, channel *model.Channel, modelName string) {
+	c.Set("original_model", modelName) // for retry
+	if channel == nil {
+		return
+	}
+	c.Set("channel", channel.Type)
+	c.Set("channel_id", channel.Id)
+	c.Set("channel_name", channel.Name)
+	ban := true
+	// parse *int to bool
+	if channel.AutoBan != nil && *channel.AutoBan == 0 {
+		ban = false
+	}
+	if nil != channel.OpenAIOrganization && "" != *channel.OpenAIOrganization {
+		c.Set("channel_organization", *channel.OpenAIOrganization)
+	}
+	c.Set("auto_ban", ban)
+	c.Set("model_mapping", channel.GetModelMapping())
+	c.Set("status_code_mapping", channel.GetStatusCodeMapping())
+	c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
+	c.Set("base_url", channel.GetBaseURL())
+	// TODO: api_version统一
+	switch channel.Type {
+	case common.ChannelTypeAzure:
+		c.Set("api_version", channel.Other)
+	case common.ChannelTypeXunfei:
+		c.Set("api_version", channel.Other)
+	//case common.ChannelTypeAIProxyLibrary:
+	//	c.Set("library_id", channel.Other)
+	case common.ChannelTypeGemini:
+		c.Set("api_version", channel.Other)
+	case common.ChannelTypeAli:
+		c.Set("plugin", channel.Other)
+	}
+}

model/ability.go

@@ -3,6 +3,8 @@ package model
 import (
 	"errors"
 	"fmt"
+	"github.com/samber/lo"
+	"gorm.io/gorm"
 	"one-api/common"
 	"strings"
 )
@@ -27,8 +29,7 @@ func GetGroupModels(group string) []string {
 	return models
 }
 
-func GetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
-	var abilities []Ability
+func getPriority(group string, model string, retry int) (int, error) {
 	groupCol := "`group`"
 	trueVal := "1"
 	if common.UsingPostgreSQL {
@@ -36,9 +37,55 @@ func GetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
 		trueVal = "true"
 	}
 
-	var err error = nil
+	var priorities []int
+	err := DB.Model(&Ability{}).
+		Select("DISTINCT(priority)").
+		Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model).
+		Order("priority DESC").              // 按优先级降序排序
+		Pluck("priority", &priorities).Error // Pluck用于将查询的结果直接扫描到一个切片中
+
+	if err != nil {
+		// 处理错误
+		return 0, err
+	}
+
+	// 确定要使用的优先级
+	var priorityToUse int
+	if retry >= len(priorities) {
+		// 如果重试次数大于优先级数，则使用最小的优先级
+		priorityToUse = priorities[len(priorities)-1]
+	} else {
+		priorityToUse = priorities[retry]
+	}
+	return priorityToUse, nil
+}
+
+func getChannelQuery(group string, model string, retry int) *gorm.DB {
+	groupCol := "`group`"
+	trueVal := "1"
+	if common.UsingPostgreSQL {
+		groupCol = `"group"`
+		trueVal = "true"
+	}
 	maxPrioritySubQuery := DB.Model(&Ability{}).Select("MAX(priority)").Where(groupCol+" = ? and model = ? and enabled = "+trueVal, group, model)
 	channelQuery := DB.Where(groupCol+" = ? and model = ? and enabled = "+trueVal+" and priority = (?)", group, model, maxPrioritySubQuery)
+	if retry != 0 {
+		priority, err := getPriority(group, model, retry)
+		if err != nil {
+			common.SysError(fmt.Sprintf("Get priority failed: %s", err.Error()))
+		} else {
+			channelQuery = DB.Where(groupCol+" = ? and model = ? and enabled = "+trueVal+" and priority = ?", group, model, priority)
+		}
+	}
+
+	return channelQuery
+}
+
+func GetRandomSatisfiedChannel(group string, model string, retry int) (*Channel, error) {
+	var abilities []Ability
+
+	var err error = nil
+	channelQuery := getChannelQuery(group, model, retry)
 	if common.UsingSQLite || common.UsingPostgreSQL {
 		err = channelQuery.Order("weight DESC").Find(&abilities).Error
 	} else {
@@ -52,21 +99,16 @@ func GetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
 		// Randomly choose one
 		weightSum := uint(0)
 		for _, ability_ := range abilities {
-			weightSum += ability_.Weight
+			weightSum += ability_.Weight + 10
 		}
-		if weightSum == 0 {
-			// All weight is 0, randomly choose one
-			channel.Id = abilities[common.GetRandomInt(len(abilities))].ChannelId
-		} else {
-			// Randomly choose one
-			weight := common.GetRandomInt(int(weightSum))
-			for _, ability_ := range abilities {
-				weight -= int(ability_.Weight)
-				//log.Printf("weight: %d, ability weight: %d", weight, *ability_.Weight)
-				if weight <= 0 {
-					channel.Id = ability_.ChannelId
-					break
-				}
+		// Randomly choose one
+		weight := common.GetRandomInt(int(weightSum))
+		for _, ability_ := range abilities {
+			weight -= int(ability_.Weight) + 10
+			//log.Printf("weight: %d, ability weight: %d", weight, *ability_.Weight)
+			if weight <= 0 {
+				channel.Id = ability_.ChannelId
+				break
 			}
 		}
 	} else {
@@ -93,7 +135,16 @@ func (channel *Channel) AddAbilities() error {
 			abilities = append(abilities, ability)
 		}
 	}
-	return DB.Create(&abilities).Error
+	if len(abilities) == 0 {
+		return nil
+	}
+	for _, chunk := range lo.Chunk(abilities, 50) {
+		err := DB.Create(&chunk).Error
+		if err != nil {
+			return err
+		}
+	}
+	return nil
 }
 
 func (channel *Channel) DeleteAbilities() error {

model/cache.go

@@ -25,9 +25,6 @@ var token2UserId = make(map[string]int)
 var token2UserIdLock sync.RWMutex
 
 func cacheSetToken(token *Token) error {
-	if !common.RedisEnabled {
-		return token.SelectUpdate()
-	}
 	jsonBytes, err := json.Marshal(token)
 	if err != nil {
 		return err
@@ -168,7 +165,11 @@ func CacheUpdateUserQuota(id int) error {
 	if err != nil {
 		return err
 	}
-	err = common.RedisSet(fmt.Sprintf("user_quota:%d", id), fmt.Sprintf("%d", quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
+	return cacheSetUserQuota(id, quota)
+}
+
+func cacheSetUserQuota(id int, quota int) error {
+	err := common.RedisSet(fmt.Sprintf("user_quota:%d", id), fmt.Sprintf("%d", quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
 	return err
 }
 
@@ -265,14 +266,14 @@ func SyncChannelCache(frequency int) {
 	}
 }
 
-func CacheGetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
+func CacheGetRandomSatisfiedChannel(group string, model string, retry int) (*Channel, error) {
 	if strings.HasPrefix(model, "gpt-4-gizmo") {
 		model = "gpt-4-gizmo-*"
 	}
 
 	// if memory cache is disabled, get channel directly from database
 	if !common.MemoryCacheEnabled {
-		return GetRandomSatisfiedChannel(group, model)
+		return GetRandomSatisfiedChannel(group, model, retry)
 	}
 	channelSyncLock.RLock()
 	defer channelSyncLock.RUnlock()
@@ -280,15 +281,27 @@ func CacheGetRandomSatisfiedChannel(group string, model string) (*Channel, error
 	if len(channels) == 0 {
 		return nil, errors.New("channel not found")
 	}
-	endIdx := len(channels)
-	// choose by priority
-	firstChannel := channels[0]
-	if firstChannel.GetPriority() > 0 {
-		for i := range channels {
-			if channels[i].GetPriority() != firstChannel.GetPriority() {
-				endIdx = i
-				break
-			}
+
+	uniquePriorities := make(map[int]bool)
+	for _, channel := range channels {
+		uniquePriorities[int(channel.GetPriority())] = true
+	}
+	var sortedUniquePriorities []int
+	for priority := range uniquePriorities {
+		sortedUniquePriorities = append(sortedUniquePriorities, priority)
+	}
+	sort.Sort(sort.Reverse(sort.IntSlice(sortedUniquePriorities)))
+
+	if retry >= len(uniquePriorities) {
+		retry = len(uniquePriorities) - 1
+	}
+	targetPriority := int64(sortedUniquePriorities[retry])
+
+	// get the priority for the given retry number
+	var targetChannels []*Channel
+	for _, channel := range channels {
+		if channel.GetPriority() == targetPriority {
+			targetChannels = append(targetChannels, channel)
 		}
 	}
 
@@ -296,20 +309,14 @@ func CacheGetRandomSatisfiedChannel(group string, model string) (*Channel, error
 	smoothingFactor := 10
 	// Calculate the total weight of all channels up to endIdx
 	totalWeight := 0
-	for _, channel := range channels[:endIdx] {
+	for _, channel := range targetChannels {
 		totalWeight += channel.GetWeight() + smoothingFactor
 	}
-
-	//if totalWeight == 0 {
-	//	// If all weights are 0, select a channel randomly
-	//	return channels[rand.Intn(endIdx)], nil
-	//}
-
 	// Generate a random value in the range [0, totalWeight)
 	randomWeight := rand.Intn(totalWeight)
 
 	// Find a channel based on its weight
-	for _, channel := range channels[:endIdx] {
+	for _, channel := range targetChannels {
 		randomWeight -= channel.GetWeight() + smoothingFactor
 		if randomWeight < 0 {
 			return channel, nil

model/channel.go

@@ -10,6 +10,7 @@ type Channel struct {
 	Type               int     `json:"type" gorm:"default:0"`
 	Key                string  `json:"key" gorm:"not null"`
 	OpenAIOrganization *string `json:"openai_organization"`
+	TestModel          *string `json:"test_model"`
 	Status             int     `json:"status" gorm:"default:1"`
 	Name               string  `json:"name" gorm:"index"`
 	Weight             *uint   `json:"weight" gorm:"default:0"`
@@ -24,8 +25,10 @@ type Channel struct {
 	Group              string  `json:"group" gorm:"type:varchar(64);default:'default'"`
 	UsedQuota          int64   `json:"used_quota" gorm:"bigint;default:0"`
 	ModelMapping       *string `json:"model_mapping" gorm:"type:varchar(1024);default:''"`
-	Priority           *int64  `json:"priority" gorm:"bigint;default:0"`
-	AutoBan            *int    `json:"auto_ban" gorm:"default:1"`
+	//MaxInputTokens     *int    `json:"max_input_tokens" gorm:"default:0"`
+	StatusCodeMapping *string `json:"status_code_mapping" gorm:"type:varchar(1024);default:''"`
+	Priority          *int64  `json:"priority" gorm:"bigint;default:0"`
+	AutoBan           *int    `json:"auto_ban" gorm:"default:1"`
 }
 
 func GetAllChannels(startIdx int, num int, selectAll bool, idSort bool) ([]*Channel, error) {
@@ -152,6 +155,13 @@ func (channel *Channel) GetModelMapping() string {
 	return *channel.ModelMapping
 }
 
+func (channel *Channel) GetStatusCodeMapping() string {
+	if channel.StatusCodeMapping == nil {
+		return ""
+	}
+	return *channel.StatusCodeMapping
+}
+
 func (channel *Channel) Insert() error {
 	var err error
 	err = DB.Create(channel).Error

model/log.go

@@ -24,6 +24,7 @@ type Log struct {
 	IsStream         bool   `json:"is_stream" gorm:"default:false"`
 	ChannelId        int    `json:"channel" gorm:"index"`
 	TokenId          int    `json:"token_id" gorm:"default:0;index"`
+	Other            string `json:"other"`
 }
 
 const (
@@ -57,12 +58,13 @@ func RecordLog(userId int, logType int, content string) {
 	}
 }
 
-func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptTokens int, completionTokens int, modelName string, tokenName string, quota int, content string, tokenId int, userQuota int, useTimeSeconds int, isStream bool) {
+func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptTokens int, completionTokens int, modelName string, tokenName string, quota int, content string, tokenId int, userQuota int, useTimeSeconds int, isStream bool, other map[string]interface{}) {
 	common.LogInfo(ctx, fmt.Sprintf("record consume log: userId=%d, 用户调用前余额=%d, channelId=%d, promptTokens=%d, completionTokens=%d, modelName=%s, tokenName=%s, quota=%d, content=%s", userId, userQuota, channelId, promptTokens, completionTokens, modelName, tokenName, quota, content))
 	if !common.LogConsumeEnabled {
 		return
 	}
 	username, _ := CacheGetUsername(userId)
+	otherStr := common.MapToJsonStr(other)
 	log := &Log{
 		UserId:           userId,
 		Username:         username,
@@ -78,6 +80,7 @@ func RecordConsumeLog(ctx context.Context, userId int, channelId int, promptToke
 		TokenId:          tokenId,
 		UseTime:          useTimeSeconds,
 		IsStream:         isStream,
+		Other:            otherStr,
 	}
 	err := DB.Create(log).Error
 	if err != nil {

model/option.go

@@ -44,12 +44,14 @@ func InitOptionMap() {
 	common.OptionMap["DataExportEnabled"] = strconv.FormatBool(common.DataExportEnabled)
 	common.OptionMap["ChannelDisableThreshold"] = strconv.FormatFloat(common.ChannelDisableThreshold, 'f', -1, 64)
 	common.OptionMap["EmailDomainRestrictionEnabled"] = strconv.FormatBool(common.EmailDomainRestrictionEnabled)
+	common.OptionMap["EmailAliasRestrictionEnabled"] = strconv.FormatBool(common.EmailAliasRestrictionEnabled)
 	common.OptionMap["EmailDomainWhitelist"] = strings.Join(common.EmailDomainWhitelist, ",")
 	common.OptionMap["SMTPServer"] = ""
 	common.OptionMap["SMTPFrom"] = ""
 	common.OptionMap["SMTPPort"] = strconv.Itoa(common.SMTPPort)
 	common.OptionMap["SMTPAccount"] = ""
 	common.OptionMap["SMTPToken"] = ""
+	common.OptionMap["SMTPSSLEnabled"] = strconv.FormatBool(common.SMTPSSLEnabled)
 	common.OptionMap["Notice"] = ""
 	common.OptionMap["About"] = ""
 	common.OptionMap["HomePageContent"] = ""
@@ -61,8 +63,8 @@ func InitOptionMap() {
 	common.OptionMap["CustomCallbackAddress"] = ""
 	common.OptionMap["EpayId"] = ""
 	common.OptionMap["EpayKey"] = ""
-	common.OptionMap["Price"] = strconv.FormatFloat(common.Price, 'f', -1, 64)
-	common.OptionMap["MinTopUp"] = strconv.Itoa(common.MinTopUp)
+	common.OptionMap["Price"] = strconv.FormatFloat(constant.Price, 'f', -1, 64)
+	common.OptionMap["MinTopUp"] = strconv.Itoa(constant.MinTopUp)
 	common.OptionMap["TopupGroupRatio"] = common.TopupGroupRatio2JSONString()
 	common.OptionMap["GitHubClientId"] = ""
 	common.OptionMap["GitHubClientSecret"] = ""
@@ -81,6 +83,7 @@ func InitOptionMap() {
 	common.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
 	common.OptionMap["ModelPrice"] = common.ModelPrice2JSONString()
 	common.OptionMap["GroupRatio"] = common.GroupRatio2JSONString()
+	common.OptionMap["CompletionRatio"] = common.CompletionRatio2JSONString()
 	common.OptionMap["TopUpLink"] = common.TopUpLink
 	common.OptionMap["ChatLink"] = common.ChatLink
 	common.OptionMap["ChatLink2"] = common.ChatLink2
@@ -90,9 +93,12 @@ func InitOptionMap() {
 	common.OptionMap["DataExportDefaultTime"] = common.DataExportDefaultTime
 	common.OptionMap["DefaultCollapseSidebar"] = strconv.FormatBool(common.DefaultCollapseSidebar)
 	common.OptionMap["MjNotifyEnabled"] = strconv.FormatBool(constant.MjNotifyEnabled)
+	common.OptionMap["MjAccountFilterEnabled"] = strconv.FormatBool(constant.MjAccountFilterEnabled)
+	common.OptionMap["MjModeClearEnabled"] = strconv.FormatBool(constant.MjModeClearEnabled)
+	common.OptionMap["MjForwardUrlEnabled"] = strconv.FormatBool(constant.MjForwardUrlEnabled)
 	common.OptionMap["CheckSensitiveEnabled"] = strconv.FormatBool(constant.CheckSensitiveEnabled)
 	common.OptionMap["CheckSensitiveOnPromptEnabled"] = strconv.FormatBool(constant.CheckSensitiveOnPromptEnabled)
-	common.OptionMap["CheckSensitiveOnCompletionEnabled"] = strconv.FormatBool(constant.CheckSensitiveOnCompletionEnabled)
+	//common.OptionMap["CheckSensitiveOnCompletionEnabled"] = strconv.FormatBool(constant.CheckSensitiveOnCompletionEnabled)
 	common.OptionMap["StopOnSensitiveEnabled"] = strconv.FormatBool(constant.StopOnSensitiveEnabled)
 	common.OptionMap["SensitiveWords"] = constant.SensitiveWordsToString()
 	common.OptionMap["StreamCacheQueueLength"] = strconv.Itoa(constant.StreamCacheQueueLength)
@@ -173,6 +179,8 @@ func updateOptionMap(key string, value string) (err error) {
 			common.RegisterEnabled = boolValue
 		case "EmailDomainRestrictionEnabled":
 			common.EmailDomainRestrictionEnabled = boolValue
+		case "EmailAliasRestrictionEnabled":
+			common.EmailAliasRestrictionEnabled = boolValue
 		case "AutomaticDisableChannelEnabled":
 			common.AutomaticDisableChannelEnabled = boolValue
 		case "AutomaticEnableChannelEnabled":
@@ -191,14 +199,22 @@ func updateOptionMap(key string, value string) (err error) {
 			common.DefaultCollapseSidebar = boolValue
 		case "MjNotifyEnabled":
 			constant.MjNotifyEnabled = boolValue
+		case "MjAccountFilterEnabled":
+			constant.MjAccountFilterEnabled = boolValue
+		case "MjModeClearEnabled":
+			constant.MjModeClearEnabled = boolValue
+		case "MjForwardUrlEnabled":
+			constant.MjForwardUrlEnabled = boolValue
 		case "CheckSensitiveEnabled":
 			constant.CheckSensitiveEnabled = boolValue
 		case "CheckSensitiveOnPromptEnabled":
 			constant.CheckSensitiveOnPromptEnabled = boolValue
-		case "CheckSensitiveOnCompletionEnabled":
-			constant.CheckSensitiveOnCompletionEnabled = boolValue
+		//case "CheckSensitiveOnCompletionEnabled":
+		//	constant.CheckSensitiveOnCompletionEnabled = boolValue
 		case "StopOnSensitiveEnabled":
 			constant.StopOnSensitiveEnabled = boolValue
+		case "SMTPSSLEnabled":
+			common.SMTPSSLEnabled = boolValue
 		}
 	}
 	switch key {
@@ -218,17 +234,17 @@ func updateOptionMap(key string, value string) (err error) {
 	case "ServerAddress":
 		common.ServerAddress = value
 	case "PayAddress":
-		common.PayAddress = value
+		constant.PayAddress = value
 	case "CustomCallbackAddress":
-		common.CustomCallbackAddress = value
+		constant.CustomCallbackAddress = value
 	case "EpayId":
-		common.EpayId = value
+		constant.EpayId = value
 	case "EpayKey":
-		common.EpayKey = value
+		constant.EpayKey = value
 	case "Price":
-		common.Price, _ = strconv.ParseFloat(value, 64)
+		constant.Price, _ = strconv.ParseFloat(value, 64)
 	case "MinTopUp":
-		common.MinTopUp, _ = strconv.Atoi(value)
+		constant.MinTopUp, _ = strconv.Atoi(value)
 	case "TopupGroupRatio":
 		err = common.UpdateTopupGroupRatioByJSONString(value)
 	case "GitHubClientId":
@@ -275,6 +291,8 @@ func updateOptionMap(key string, value string) (err error) {
 		err = common.UpdateModelRatioByJSONString(value)
 	case "GroupRatio":
 		err = common.UpdateGroupRatioByJSONString(value)
+	case "CompletionRatio":
+		err = common.UpdateCompletionRatioByJSONString(value)
 	case "ModelPrice":
 		err = common.UpdateModelPriceByJSONString(value)
 	case "TopUpLink":

model/redemption.go

@@ -56,7 +56,7 @@ func Redeem(key string, userId int) (quota int, err error) {
 	if common.UsingPostgreSQL {
 		keyCol = `"key"`
 	}
-
+	common.RandomSleep()
 	err = DB.Transaction(func(tx *gorm.DB) error {
 		err := tx.Set("gorm:query_option", "FOR UPDATE").Where(keyCol+" = ?", key).First(redemption).Error
 		if err != nil {

model/token.go

@@ -102,6 +102,11 @@ func GetTokenById(id int) (*Token, error) {
 	token := Token{Id: id}
 	var err error = nil
 	err = DB.First(&token, "id = ?", id).Error
+	if err != nil {
+		if common.RedisEnabled {
+			go cacheSetToken(&token)
+		}
+	}
 	return &token, err
 }
 

model/user.go

@@ -4,6 +4,7 @@ import (
 	"errors"
 	"fmt"
 	"one-api/common"
+	"strconv"
 	"strings"
 	"time"
 
@@ -72,8 +73,35 @@ func GetAllUsers(startIdx int, num int) (users []*User, err error) {
 	return users, err
 }
 
-func SearchUsers(keyword string) (users []*User, err error) {
-	err = DB.Omit("password").Where("id = ? or username LIKE ? or email LIKE ? or display_name LIKE ?", keyword, keyword+"%", keyword+"%", keyword+"%").Find(&users).Error
+func SearchUsers(keyword string, group string) ([]*User, error) {
+	var users []*User
+	var err error
+
+	// 尝试将关键字转换为整数ID
+	keywordInt, err := strconv.Atoi(keyword)
+	if err == nil {
+		// 如果转换成功，按照ID和可选的组别搜索用户
+		query := DB.Unscoped().Omit("password").Where("`id` = ?", keywordInt)
+		if group != "" {
+			query = query.Where("`group` = ?", group) // 使用反引号包围group
+		}
+		err = query.Find(&users).Error
+		if err != nil || len(users) > 0 {
+			return users, err
+		}
+	}
+
+	err = nil
+
+	query := DB.Unscoped().Omit("password")
+	likeCondition := "`username` LIKE ? OR `email` LIKE ? OR `display_name` LIKE ?"
+	if group != "" {
+		query = query.Where("("+likeCondition+") AND `group` = ?", "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%", group)
+	} else {
+		query = query.Where(likeCondition, "%"+keyword+"%", "%"+keyword+"%", "%"+keyword+"%")
+	}
+	err = query.Find(&users).Error
+
 	return users, err
 }
 
@@ -210,6 +238,36 @@ func (user *User) Update(updatePassword bool) error {
 	if err == nil {
 		if common.RedisEnabled {
 			_ = common.RedisSet(fmt.Sprintf("user_group:%d", user.Id), user.Group, time.Duration(UserId2GroupCacheSeconds)*time.Second)
+			_ = common.RedisSet(fmt.Sprintf("user_quota:%d", user.Id), strconv.Itoa(user.Quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
+		}
+	}
+	return err
+}
+
+func (user *User) Edit(updatePassword bool) error {
+	var err error
+	if updatePassword {
+		user.Password, err = common.Password2Hash(user.Password)
+		if err != nil {
+			return err
+		}
+	}
+	newUser := *user
+	updates := map[string]interface{}{
+		"username":     newUser.Username,
+		"display_name": newUser.DisplayName,
+		"group":        newUser.Group,
+		"quota":        newUser.Quota,
+	}
+	if updatePassword {
+		updates["password"] = newUser.Password
+	}
+	DB.First(&user, user.Id)
+	err = DB.Model(user).Updates(updates).Error
+	if err == nil {
+		if common.RedisEnabled {
+			_ = common.RedisSet(fmt.Sprintf("user_group:%d", user.Id), user.Group, time.Duration(UserId2GroupCacheSeconds)*time.Second)
+			_ = common.RedisSet(fmt.Sprintf("user_quota:%d", user.Id), strconv.Itoa(user.Quota), time.Duration(UserId2QuotaCacheSeconds)*time.Second)
 		}
 	}
 	return err
@@ -370,6 +428,11 @@ func ValidateAccessToken(token string) (user *User) {
 
 func GetUserQuota(id int) (quota int, err error) {
 	err = DB.Model(&User{}).Where("id = ?", id).Select("quota").Find(&quota).Error
+	if err != nil {
+		if common.RedisEnabled {
+			go cacheSetUserQuota(id, quota)
+		}
+	}
 	return quota, err
 }
 

relay/channel/adapter.go

@@ -15,7 +15,7 @@ type Adaptor interface {
 	SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error
 	ConvertRequest(c *gin.Context, relayMode int, request *dto.GeneralOpenAIRequest) (any, error)
 	DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (*http.Response, error)
-	DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse)
+	DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode)
 	GetModelList() []string
 	GetChannelName() string
 }

relay/channel/ai360/constants.go

@@ -6,3 +6,5 @@ var ModelList = []string{
 	"embedding_s1_v1",
 	"semantic_similarity_s1_v1",
 }
+
+var ChannelName = "ai360"

relay/channel/ali/adaptor.go

@@ -57,7 +57,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		err, usage = aliStreamHandler(c, resp)
 	} else {

relay/channel/ali/relay-ali.go

@@ -136,7 +136,7 @@ func responseAli2OpenAI(response *AliChatResponse) *dto.OpenAITextResponse {
 
 func streamResponseAli2OpenAI(aliResponse *AliChatResponse) *dto.ChatCompletionsStreamResponse {
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = aliResponse.Output.Text
+	choice.Delta.SetContentString(aliResponse.Output.Text)
 	if aliResponse.Output.FinishReason != "null" {
 		finishReason := aliResponse.Output.FinishReason
 		choice.FinishReason = &finishReason
@@ -199,7 +199,7 @@ func aliStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIErrorWith
 				usage.TotalTokens = aliResponse.Usage.InputTokens + aliResponse.Usage.OutputTokens
 			}
 			response := streamResponseAli2OpenAI(&aliResponse)
-			response.Choices[0].Delta.Content = strings.TrimPrefix(response.Choices[0].Delta.Content, lastResponseText)
+			response.Choices[0].Delta.SetContentString(strings.TrimPrefix(response.Choices[0].Delta.GetContentString(), lastResponseText))
 			lastResponseText = aliResponse.Output.Text
 			jsonResponse, err := json.Marshal(response)
 			if err != nil {

relay/channel/aws/adaptor.go

@@ -0,0 +1,79 @@
+package aws
+
+import (
+	"errors"
+	"github.com/gin-gonic/gin"
+	"io"
+	"net/http"
+	"one-api/dto"
+	"one-api/relay/channel/claude"
+	relaycommon "one-api/relay/common"
+	"strings"
+)
+
+const (
+	RequestModeCompletion = 1
+	RequestModeMessage    = 2
+)
+
+type Adaptor struct {
+	RequestMode int
+}
+
+func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) {
+	if strings.HasPrefix(info.UpstreamModelName, "claude-3") {
+		a.RequestMode = RequestModeMessage
+	} else {
+		a.RequestMode = RequestModeCompletion
+	}
+}
+
+func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
+	return "", nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *dto.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+
+	var claudeReq *claude.ClaudeRequest
+	var err error
+	if a.RequestMode == RequestModeCompletion {
+		claudeReq = claude.RequestOpenAI2ClaudeComplete(*request)
+	} else {
+		claudeReq, err = claude.RequestOpenAI2ClaudeMessage(*request)
+	}
+	c.Set("request_model", request.Model)
+	c.Set("converted_request", claudeReq)
+	return claudeReq, err
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (*http.Response, error) {
+	return nil, nil
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
+	if info.IsStream {
+		err, usage = awsStreamHandler(c, info, a.RequestMode)
+	} else {
+		err, usage = awsHandler(c, info, a.RequestMode)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() (models []string) {
+	for n := range awsModelIDMap {
+		models = append(models, n)
+	}
+
+	return
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return ChannelName
+}

relay/channel/aws/constants.go

@@ -0,0 +1,12 @@
+package aws
+
+var awsModelIDMap = map[string]string{
+	"claude-instant-1.2":       "anthropic.claude-instant-v1",
+	"claude-2.0":               "anthropic.claude-v2",
+	"claude-2.1":               "anthropic.claude-v2:1",
+	"claude-3-sonnet-20240229": "anthropic.claude-3-sonnet-20240229-v1:0",
+	"claude-3-opus-20240229":   "anthropic.claude-3-opus-20240229-v1:0",
+	"claude-3-haiku-20240307":  "anthropic.claude-3-haiku-20240307-v1:0",
+}
+
+var ChannelName = "aws"

relay/channel/aws/dto.go

@@ -0,0 +1,15 @@
+package aws
+
+import "one-api/relay/channel/claude"
+
+type AwsClaudeRequest struct {
+	// AnthropicVersion should be "bedrock-2023-05-31"
+	AnthropicVersion string                 `json:"anthropic_version"`
+	System           string                 `json:"system"`
+	Messages         []claude.ClaudeMessage `json:"messages"`
+	MaxTokens        int                    `json:"max_tokens,omitempty"`
+	Temperature      float64                `json:"temperature,omitempty"`
+	TopP             float64                `json:"top_p,omitempty"`
+	TopK             int                    `json:"top_k,omitempty"`
+	StopSequences    []string               `json:"stop_sequences,omitempty"`
+}

relay/channel/aws/relay-aws.go

@@ -0,0 +1,213 @@
+package aws
+
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/jinzhu/copier"
+	"github.com/pkg/errors"
+	"io"
+	"net/http"
+	"one-api/common"
+	relaymodel "one-api/dto"
+	"one-api/relay/channel/claude"
+	relaycommon "one-api/relay/common"
+	"strings"
+
+	"github.com/aws/aws-sdk-go-v2/aws"
+	"github.com/aws/aws-sdk-go-v2/credentials"
+	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime"
+	"github.com/aws/aws-sdk-go-v2/service/bedrockruntime/types"
+)
+
+func newAwsClient(c *gin.Context, info *relaycommon.RelayInfo) (*bedrockruntime.Client, error) {
+	awsSecret := strings.Split(info.ApiKey, "|")
+	if len(awsSecret) != 3 {
+		return nil, errors.New("invalid aws secret key")
+	}
+	ak := awsSecret[0]
+	sk := awsSecret[1]
+	region := awsSecret[2]
+	client := bedrockruntime.New(bedrockruntime.Options{
+		Region:      region,
+		Credentials: aws.NewCredentialsCache(credentials.NewStaticCredentialsProvider(ak, sk, "")),
+	})
+
+	return client, nil
+}
+
+func wrapErr(err error) *relaymodel.OpenAIErrorWithStatusCode {
+	return &relaymodel.OpenAIErrorWithStatusCode{
+		StatusCode: http.StatusInternalServerError,
+		Error: relaymodel.OpenAIError{
+			Message: fmt.Sprintf("%s", err.Error()),
+		},
+	}
+}
+
+func awsModelID(requestModel string) (string, error) {
+	if awsModelID, ok := awsModelIDMap[requestModel]; ok {
+		return awsModelID, nil
+	}
+
+	return "", errors.Errorf("model %s not found", requestModel)
+}
+
+func awsHandler(c *gin.Context, info *relaycommon.RelayInfo, requestMode int) (*relaymodel.OpenAIErrorWithStatusCode, *relaymodel.Usage) {
+	awsCli, err := newAwsClient(c, info)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
+	}
+
+	awsModelId, err := awsModelID(c.GetString("request_model"))
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "awsModelID")), nil
+	}
+
+	awsReq := &bedrockruntime.InvokeModelInput{
+		ModelId:     aws.String(awsModelId),
+		Accept:      aws.String("application/json"),
+		ContentType: aws.String("application/json"),
+	}
+
+	claudeReq_, ok := c.Get("converted_request")
+	if !ok {
+		return wrapErr(errors.New("request not found")), nil
+	}
+	claudeReq := claudeReq_.(*claude.ClaudeRequest)
+	awsClaudeReq := &AwsClaudeRequest{
+		AnthropicVersion: "bedrock-2023-05-31",
+	}
+	if err = copier.Copy(awsClaudeReq, claudeReq); err != nil {
+		return wrapErr(errors.Wrap(err, "copy request")), nil
+	}
+
+	awsReq.Body, err = json.Marshal(awsClaudeReq)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "marshal request")), nil
+	}
+
+	awsResp, err := awsCli.InvokeModel(c.Request.Context(), awsReq)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "InvokeModel")), nil
+	}
+
+	claudeResponse := new(claude.ClaudeResponse)
+	err = json.Unmarshal(awsResp.Body, claudeResponse)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "unmarshal response")), nil
+	}
+
+	openaiResp := claude.ResponseClaude2OpenAI(requestMode, claudeResponse)
+	usage := relaymodel.Usage{
+		PromptTokens:     claudeResponse.Usage.InputTokens,
+		CompletionTokens: claudeResponse.Usage.OutputTokens,
+		TotalTokens:      claudeResponse.Usage.InputTokens + claudeResponse.Usage.OutputTokens,
+	}
+	openaiResp.Usage = usage
+
+	c.JSON(http.StatusOK, openaiResp)
+	return nil, &usage
+}
+
+func awsStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, requestMode int) (*relaymodel.OpenAIErrorWithStatusCode, *relaymodel.Usage) {
+	awsCli, err := newAwsClient(c, info)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "newAwsClient")), nil
+	}
+
+	awsModelId, err := awsModelID(c.GetString("request_model"))
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "awsModelID")), nil
+	}
+
+	awsReq := &bedrockruntime.InvokeModelWithResponseStreamInput{
+		ModelId:     aws.String(awsModelId),
+		Accept:      aws.String("application/json"),
+		ContentType: aws.String("application/json"),
+	}
+
+	claudeReq_, ok := c.Get("converted_request")
+	if !ok {
+		return wrapErr(errors.New("request not found")), nil
+	}
+	claudeReq := claudeReq_.(*claude.ClaudeRequest)
+
+	awsClaudeReq := &AwsClaudeRequest{
+		AnthropicVersion: "bedrock-2023-05-31",
+	}
+	if err = copier.Copy(awsClaudeReq, claudeReq); err != nil {
+		return wrapErr(errors.Wrap(err, "copy request")), nil
+	}
+	awsReq.Body, err = json.Marshal(awsClaudeReq)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "marshal request")), nil
+	}
+
+	awsResp, err := awsCli.InvokeModelWithResponseStream(c.Request.Context(), awsReq)
+	if err != nil {
+		return wrapErr(errors.Wrap(err, "InvokeModelWithResponseStream")), nil
+	}
+	stream := awsResp.GetStream()
+	defer stream.Close()
+
+	c.Writer.Header().Set("Content-Type", "text/event-stream")
+	var usage relaymodel.Usage
+	var id string
+	var model string
+	createdTime := common.GetTimestamp()
+	c.Stream(func(w io.Writer) bool {
+		event, ok := <-stream.Events()
+		if !ok {
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+
+		switch v := event.(type) {
+		case *types.ResponseStreamMemberChunk:
+			claudeResp := new(claude.ClaudeResponse)
+			err := json.NewDecoder(bytes.NewReader(v.Value.Bytes)).Decode(claudeResp)
+			if err != nil {
+				common.SysError("error unmarshalling stream response: " + err.Error())
+				return false
+			}
+
+			response, claudeUsage := claude.StreamResponseClaude2OpenAI(requestMode, claudeResp)
+			if claudeUsage != nil {
+				usage.PromptTokens += claudeUsage.InputTokens
+				usage.CompletionTokens += claudeUsage.OutputTokens
+			}
+
+			if response == nil {
+				return true
+			}
+
+			if response.Id != "" {
+				id = response.Id
+			}
+			if response.Model != "" {
+				model = response.Model
+			}
+			response.Created = createdTime
+			response.Id = id
+			response.Model = model
+
+			jsonStr, err := json.Marshal(response)
+			if err != nil {
+				common.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonStr)})
+			return true
+		case *types.UnknownUnionMember:
+			fmt.Println("unknown tag:", v.Tag)
+			return false
+		default:
+			fmt.Println("union is nil or unknown type")
+			return false
+		}
+	})
+
+	return nil, &usage
+}

relay/channel/baidu/adaptor.go

@@ -69,7 +69,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		err, usage = baiduStreamHandler(c, resp)
 	} else {

relay/channel/baidu/relay-baidu.go

@@ -57,7 +57,7 @@ func responseBaidu2OpenAI(response *BaiduChatResponse) *dto.OpenAITextResponse {
 
 func streamResponseBaidu2OpenAI(baiduResponse *BaiduChatStreamResponse) *dto.ChatCompletionsStreamResponse {
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = baiduResponse.Result
+	choice.Delta.SetContentString(baiduResponse.Result)
 	if baiduResponse.IsEnd {
 		choice.FinishReason = &relaycommon.StopFinishReason
 	}

relay/channel/claude/adaptor.go

@@ -53,9 +53,9 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *dto.Gen
 		return nil, errors.New("request is nil")
 	}
 	if a.RequestMode == RequestModeCompletion {
-		return requestOpenAI2ClaudeComplete(*request), nil
+		return RequestOpenAI2ClaudeComplete(*request), nil
 	} else {
-		return requestOpenAI2ClaudeMessage(*request)
+		return RequestOpenAI2ClaudeMessage(*request)
 	}
 }
 
@@ -63,7 +63,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		err, usage = claudeStreamHandler(a.RequestMode, info.UpstreamModelName, info.PromptTokens, c, resp)
 	} else {

relay/channel/claude/dto.go

@@ -28,8 +28,8 @@ type ClaudeRequest struct {
 	Prompt            string          `json:"prompt,omitempty"`
 	System            string          `json:"system,omitempty"`
 	Messages          []ClaudeMessage `json:"messages,omitempty"`
-	MaxTokensToSample uint            `json:"max_tokens_to_sample,omitempty"`
 	MaxTokens         uint            `json:"max_tokens,omitempty"`
+	MaxTokensToSample uint            `json:"max_tokens_to_sample,omitempty"`
 	StopSequences     []string        `json:"stop_sequences,omitempty"`
 	Temperature       float64         `json:"temperature,omitempty"`
 	TopP              float64         `json:"top_p,omitempty"`

relay/channel/claude/relay-claude.go

@@ -8,7 +8,6 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
-	"one-api/constant"
 	"one-api/dto"
 	"one-api/service"
 	"strings"
@@ -21,24 +20,24 @@ func stopReasonClaude2OpenAI(reason string) string {
 	case "end_turn":
 		return "stop"
 	case "max_tokens":
-		return "length"
+		return "max_tokens"
 	default:
 		return reason
 	}
 }
 
-func requestOpenAI2ClaudeComplete(textRequest dto.GeneralOpenAIRequest) *ClaudeRequest {
+func RequestOpenAI2ClaudeComplete(textRequest dto.GeneralOpenAIRequest) *ClaudeRequest {
 	claudeRequest := ClaudeRequest{
-		Model:             textRequest.Model,
-		Prompt:            "",
-		MaxTokensToSample: textRequest.MaxTokens,
-		StopSequences:     nil,
-		Temperature:       textRequest.Temperature,
-		TopP:              textRequest.TopP,
-		Stream:            textRequest.Stream,
+		Model:         textRequest.Model,
+		Prompt:        "",
+		StopSequences: nil,
+		Temperature:   textRequest.Temperature,
+		TopP:          textRequest.TopP,
+		TopK:          textRequest.TopK,
+		Stream:        textRequest.Stream,
 	}
 	if claudeRequest.MaxTokensToSample == 0 {
-		claudeRequest.MaxTokensToSample = 1000000
+		claudeRequest.MaxTokensToSample = 4096
 	}
 	prompt := ""
 	for _, message := range textRequest.Messages {
@@ -57,22 +56,65 @@ func requestOpenAI2ClaudeComplete(textRequest dto.GeneralOpenAIRequest) *ClaudeR
 	return &claudeRequest
 }
 
-func requestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*ClaudeRequest, error) {
+func RequestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*ClaudeRequest, error) {
 	claudeRequest := ClaudeRequest{
 		Model:         textRequest.Model,
 		MaxTokens:     textRequest.MaxTokens,
 		StopSequences: nil,
 		Temperature:   textRequest.Temperature,
 		TopP:          textRequest.TopP,
+		TopK:          textRequest.TopK,
 		Stream:        textRequest.Stream,
 	}
 	if claudeRequest.MaxTokens == 0 {
 		claudeRequest.MaxTokens = 4096
 	}
+	formatMessages := make([]dto.Message, 0)
+	var lastMessage *dto.Message
+	for i, message := range textRequest.Messages {
+		//if message.Role == "system" {
+		//	if i != 0 {
+		//		message.Role = "user"
+		//	}
+		//}
+		if message.Role == "" {
+			textRequest.Messages[i].Role = "user"
+		}
+		fmtMessage := dto.Message{
+			Role:    message.Role,
+			Content: message.Content,
+		}
+		if lastMessage != nil && lastMessage.Role == message.Role {
+			if lastMessage.IsStringContent() && message.IsStringContent() {
+				content, _ := json.Marshal(strings.Trim(fmt.Sprintf("%s %s", lastMessage.StringContent(), message.StringContent()), "\""))
+				fmtMessage.Content = content
+				// delete last message
+				formatMessages = formatMessages[:len(formatMessages)-1]
+			}
+		}
+		if fmtMessage.Content == nil {
+			content, _ := json.Marshal("...")
+			fmtMessage.Content = content
+		}
+		formatMessages = append(formatMessages, fmtMessage)
+		lastMessage = &textRequest.Messages[i]
+	}
+
 	claudeMessages := make([]ClaudeMessage, 0)
-	for _, message := range textRequest.Messages {
+	for _, message := range formatMessages {
 		if message.Role == "system" {
-			claudeRequest.System = message.StringContent()
+			if message.IsStringContent() {
+				claudeRequest.System = message.StringContent()
+			} else {
+				contents := message.ParseContent()
+				content := ""
+				for _, ctx := range contents {
+					if ctx.Type == "text" {
+						content += ctx.Text
+					}
+				}
+				claudeRequest.System = content
+			}
 		} else {
 			claudeMessage := ClaudeMessage{
 				Role: message.Role,
@@ -117,11 +159,10 @@ func requestOpenAI2ClaudeMessage(textRequest dto.GeneralOpenAIRequest) (*ClaudeR
 	}
 	claudeRequest.Prompt = ""
 	claudeRequest.Messages = claudeMessages
-
 	return &claudeRequest, nil
 }
 
-func streamResponseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) (*dto.ChatCompletionsStreamResponse, *ClaudeUsage) {
+func StreamResponseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) (*dto.ChatCompletionsStreamResponse, *ClaudeUsage) {
 	var response dto.ChatCompletionsStreamResponse
 	var claudeUsage *ClaudeUsage
 	response.Object = "chat.completion.chunk"
@@ -129,7 +170,7 @@ func streamResponseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) (*
 	response.Choices = make([]dto.ChatCompletionsStreamResponseChoice, 0)
 	var choice dto.ChatCompletionsStreamResponseChoice
 	if reqMode == RequestModeCompletion {
-		choice.Delta.Content = claudeResponse.Completion
+		choice.Delta.SetContentString(claudeResponse.Completion)
 		finishReason := stopReasonClaude2OpenAI(claudeResponse.StopReason)
 		if finishReason != "null" {
 			choice.FinishReason = &finishReason
@@ -139,25 +180,34 @@ func streamResponseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) (*
 			response.Id = claudeResponse.Message.Id
 			response.Model = claudeResponse.Message.Model
 			claudeUsage = &claudeResponse.Message.Usage
+			choice.Delta.SetContentString("")
+			choice.Delta.Role = "assistant"
+		} else if claudeResponse.Type == "content_block_start" {
+			return nil, nil
 		} else if claudeResponse.Type == "content_block_delta" {
 			choice.Index = claudeResponse.Index
-			choice.Delta.Content = claudeResponse.Delta.Text
+			choice.Delta.SetContentString(claudeResponse.Delta.Text)
 		} else if claudeResponse.Type == "message_delta" {
 			finishReason := stopReasonClaude2OpenAI(*claudeResponse.Delta.StopReason)
 			if finishReason != "null" {
 				choice.FinishReason = &finishReason
 			}
 			claudeUsage = &claudeResponse.Usage
+		} else if claudeResponse.Type == "message_stop" {
+			return nil, nil
+		} else {
+			return nil, nil
 		}
 	}
 	if claudeUsage == nil {
 		claudeUsage = &ClaudeUsage{}
 	}
 	response.Choices = append(response.Choices, choice)
+
 	return &response, claudeUsage
 }
 
-func responseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) *dto.OpenAITextResponse {
+func ResponseClaude2OpenAI(reqMode int, claudeResponse *ClaudeResponse) *dto.OpenAITextResponse {
 	choices := make([]dto.OpenAITextResponseChoice, 0)
 	fullTextResponse := dto.OpenAITextResponse{
 		Id:      fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
@@ -241,7 +291,10 @@ func claudeStreamHandler(requestMode int, modelName string, promptTokens int, c
 				return true
 			}
 
-			response, claudeUsage := streamResponseClaude2OpenAI(requestMode, &claudeResponse)
+			response, claudeUsage := StreamResponseClaude2OpenAI(requestMode, &claudeResponse)
+			if response == nil {
+				return true
+			}
 			if requestMode == RequestModeCompletion {
 				responseText += claudeResponse.Completion
 				responseId = response.Id
@@ -316,8 +369,8 @@ func claudeHandler(requestMode int, c *gin.Context, resp *http.Response, promptT
 			StatusCode: resp.StatusCode,
 		}, nil
 	}
-	fullTextResponse := responseClaude2OpenAI(requestMode, &claudeResponse)
-	completionTokens, err, _ := service.CountTokenText(claudeResponse.Completion, model, constant.ShouldCheckCompletionSensitive())
+	fullTextResponse := ResponseClaude2OpenAI(requestMode, &claudeResponse)
+	completionTokens, err, _ := service.CountTokenText(claudeResponse.Completion, model, false)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "count_token_text_failed", http.StatusInternalServerError), nil
 	}

relay/channel/cohere/adaptor.go

@@ -0,0 +1,52 @@
+package cohere
+
+import (
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"io"
+	"net/http"
+	"one-api/dto"
+	"one-api/relay/channel"
+	relaycommon "one-api/relay/common"
+)
+
+type Adaptor struct {
+}
+
+func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) {
+}
+
+func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
+	return fmt.Sprintf("%s/v1/chat", info.BaseUrl), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
+	channel.SetupApiRequestHeader(info, c, req)
+	req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", info.ApiKey))
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *dto.GeneralOpenAIRequest) (any, error) {
+	return requestOpenAI2Cohere(*request), nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoApiRequest(a, c, info, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
+	if info.IsStream {
+		err, usage = cohereStreamHandler(c, resp, info.UpstreamModelName, info.PromptTokens)
+	} else {
+		err, usage = cohereHandler(c, resp, info.UpstreamModelName, info.PromptTokens)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return ChannelName
+}

relay/channel/cohere/constant.go

@@ -0,0 +1,7 @@
+package cohere
+
+var ModelList = []string{
+	"command-r", "command-r-plus", "command-light", "command-light-nightly", "command", "command-nightly",
+}
+
+var ChannelName = "cohere"

relay/channel/cohere/dto.go

@@ -0,0 +1,44 @@
+package cohere
+
+type CohereRequest struct {
+	Model       string        `json:"model"`
+	ChatHistory []ChatHistory `json:"chat_history"`
+	Message     string        `json:"message"`
+	Stream      bool          `json:"stream"`
+	MaxTokens   int64         `json:"max_tokens"`
+}
+
+type ChatHistory struct {
+	Role    string `json:"role"`
+	Message string `json:"message"`
+}
+
+type CohereResponse struct {
+	IsFinished   bool                  `json:"is_finished"`
+	EventType    string                `json:"event_type"`
+	Text         string                `json:"text,omitempty"`
+	FinishReason string                `json:"finish_reason,omitempty"`
+	Response     *CohereResponseResult `json:"response"`
+}
+
+type CohereResponseResult struct {
+	ResponseId   string     `json:"response_id"`
+	FinishReason string     `json:"finish_reason,omitempty"`
+	Text         string     `json:"text"`
+	Meta         CohereMeta `json:"meta"`
+}
+
+type CohereMeta struct {
+	//Tokens CohereTokens `json:"tokens"`
+	BilledUnits CohereBilledUnits `json:"billed_units"`
+}
+
+type CohereBilledUnits struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
+}
+
+type CohereTokens struct {
+	InputTokens  int `json:"input_tokens"`
+	OutputTokens int `json:"output_tokens"`
+}

relay/channel/cohere/relay-cohere.go

@@ -0,0 +1,189 @@
+package cohere
+
+import (
+	"bufio"
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"io"
+	"net/http"
+	"one-api/common"
+	"one-api/dto"
+	"one-api/service"
+	"strings"
+)
+
+func requestOpenAI2Cohere(textRequest dto.GeneralOpenAIRequest) *CohereRequest {
+	cohereReq := CohereRequest{
+		Model:       textRequest.Model,
+		ChatHistory: []ChatHistory{},
+		Message:     "",
+		Stream:      textRequest.Stream,
+		MaxTokens:   textRequest.GetMaxTokens(),
+	}
+	if cohereReq.MaxTokens == 0 {
+		cohereReq.MaxTokens = 4000
+	}
+	for _, msg := range textRequest.Messages {
+		if msg.Role == "user" {
+			cohereReq.Message = msg.StringContent()
+		} else {
+			var role string
+			if msg.Role == "assistant" {
+				role = "CHATBOT"
+			} else if msg.Role == "system" {
+				role = "SYSTEM"
+			} else {
+				role = "USER"
+			}
+			cohereReq.ChatHistory = append(cohereReq.ChatHistory, ChatHistory{
+				Role:    role,
+				Message: msg.StringContent(),
+			})
+		}
+	}
+	return &cohereReq
+}
+
+func stopReasonCohere2OpenAI(reason string) string {
+	switch reason {
+	case "COMPLETE":
+		return "stop"
+	case "MAX_TOKENS":
+		return "max_tokens"
+	default:
+		return reason
+	}
+}
+
+func cohereStreamHandler(c *gin.Context, resp *http.Response, modelName string, promptTokens int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	responseId := fmt.Sprintf("chatcmpl-%s", common.GetUUID())
+	createdTime := common.GetTimestamp()
+	usage := &dto.Usage{}
+	responseText := ""
+	scanner := bufio.NewScanner(resp.Body)
+	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
+		if atEOF && len(data) == 0 {
+			return 0, nil, nil
+		}
+		if i := strings.Index(string(data), "\n"); i >= 0 {
+			return i + 1, data[0:i], nil
+		}
+		if atEOF {
+			return len(data), data, nil
+		}
+		return 0, nil, nil
+	})
+	dataChan := make(chan string)
+	stopChan := make(chan bool)
+	go func() {
+		for scanner.Scan() {
+			data := scanner.Text()
+			dataChan <- data
+		}
+		stopChan <- true
+	}()
+	service.SetEventStreamHeaders(c)
+	c.Stream(func(w io.Writer) bool {
+		select {
+		case data := <-dataChan:
+			data = strings.TrimSuffix(data, "\r")
+			var cohereResp CohereResponse
+			err := json.Unmarshal([]byte(data), &cohereResp)
+			if err != nil {
+				common.SysError("error unmarshalling stream response: " + err.Error())
+				return true
+			}
+			var openaiResp dto.ChatCompletionsStreamResponse
+			openaiResp.Id = responseId
+			openaiResp.Created = createdTime
+			openaiResp.Object = "chat.completion.chunk"
+			openaiResp.Model = modelName
+			if cohereResp.IsFinished {
+				finishReason := stopReasonCohere2OpenAI(cohereResp.FinishReason)
+				openaiResp.Choices = []dto.ChatCompletionsStreamResponseChoice{
+					{
+						Delta:        dto.ChatCompletionsStreamResponseChoiceDelta{},
+						Index:        0,
+						FinishReason: &finishReason,
+					},
+				}
+				if cohereResp.Response != nil {
+					usage.PromptTokens = cohereResp.Response.Meta.BilledUnits.InputTokens
+					usage.CompletionTokens = cohereResp.Response.Meta.BilledUnits.OutputTokens
+				}
+			} else {
+				openaiResp.Choices = []dto.ChatCompletionsStreamResponseChoice{
+					{
+						Delta: dto.ChatCompletionsStreamResponseChoiceDelta{
+							Role:    "assistant",
+							Content: &cohereResp.Text,
+						},
+						Index: 0,
+					},
+				}
+				responseText += cohereResp.Text
+			}
+			jsonStr, err := json.Marshal(openaiResp)
+			if err != nil {
+				common.SysError("error marshalling stream response: " + err.Error())
+				return true
+			}
+			c.Render(-1, common.CustomEvent{Data: "data: " + string(jsonStr)})
+			return true
+		case <-stopChan:
+			c.Render(-1, common.CustomEvent{Data: "data: [DONE]"})
+			return false
+		}
+	})
+	if usage.PromptTokens == 0 {
+		usage, _ = service.ResponseText2Usage(responseText, modelName, promptTokens)
+	}
+	return nil, usage
+}
+
+func cohereHandler(c *gin.Context, resp *http.Response, modelName string, promptTokens int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	createdTime := common.GetTimestamp()
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	var cohereResp CohereResponseResult
+	err = json.Unmarshal(responseBody, &cohereResp)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	usage := dto.Usage{}
+	usage.PromptTokens = cohereResp.Meta.BilledUnits.InputTokens
+	usage.CompletionTokens = cohereResp.Meta.BilledUnits.OutputTokens
+	usage.TotalTokens = cohereResp.Meta.BilledUnits.InputTokens + cohereResp.Meta.BilledUnits.OutputTokens
+
+	var openaiResp dto.TextResponse
+	openaiResp.Id = cohereResp.ResponseId
+	openaiResp.Created = createdTime
+	openaiResp.Object = "chat.completion"
+	openaiResp.Model = modelName
+	openaiResp.Usage = usage
+
+	content, _ := json.Marshal(cohereResp.Text)
+	openaiResp.Choices = []dto.OpenAITextResponseChoice{
+		{
+			Index:        0,
+			Message:      dto.Message{Content: content, Role: "assistant"},
+			FinishReason: stopReasonCohere2OpenAI(cohereResp.FinishReason),
+		},
+	}
+
+	jsonResponse, err := json.Marshal(openaiResp)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &usage
+}

relay/channel/gemini/adaptor.go

@@ -18,16 +18,28 @@ type Adaptor struct {
 func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIRequest) {
 }
 
+// 定义一个映射，存储模型名称和对应的版本
+var modelVersionMap = map[string]string{
+    "gemini-1.5-pro-latest": "v1beta",
+    "gemini-ultra":   "v1beta",
+}
+
 func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
-	version := "v1"
-	if info.ApiVersion != "" {
-		version = info.ApiVersion
-	}
-	action := "generateContent"
-	if info.IsStream {
-		action = "streamGenerateContent"
-	}
-	return fmt.Sprintf("%s/%s/models/%s:%s", info.BaseUrl, version, info.UpstreamModelName, action), nil
+    // 从映射中获取模型名称对应的版本，如果找不到就使用 info.ApiVersion 或默认的版本 "v1"
+    version, beta := modelVersionMap[info.UpstreamModelName]
+    if !beta {
+        if info.ApiVersion != "" {
+            version = info.ApiVersion
+        } else {
+            version = "v1"
+        }
+    }
+
+    action := "generateContent"
+    if info.IsStream {
+        action = "streamGenerateContent"
+    }
+    return fmt.Sprintf("%s/%s/models/%s:%s", info.BaseUrl, version, info.UpstreamModelName, action), nil
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
@@ -47,7 +59,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
 		err, responseText = geminiChatStreamHandler(c, resp)

relay/channel/gemini/constant.go

@@ -5,8 +5,8 @@ const (
 )
 
 var ModelList = []string{
-	"gemini-pro",
-	"gemini-pro-vision",
+	"gemini-1.0-pro-latest", "gemini-1.0-pro-001", "gemini-1.5-pro-latest", "gemini-ultra",
+	"gemini-1.0-pro-vision-latest", "gemini-1.0-pro-vision-001",
 }
 
 var ChannelName = "google gemini"

relay/channel/gemini/relay-gemini.go

@@ -7,7 +7,6 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
-	"one-api/constant"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
@@ -152,7 +151,7 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *dto.OpenAITextResp
 
 func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) *dto.ChatCompletionsStreamResponse {
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = geminiResponse.GetResponseText()
+	choice.Delta.SetContentString(geminiResponse.GetResponseText())
 	choice.FinishReason = &relaycommon.StopFinishReason
 	var response dto.ChatCompletionsStreamResponse
 	response.Object = "chat.completion.chunk"
@@ -204,7 +203,7 @@ func geminiChatStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIEr
 			err := json.Unmarshal([]byte(data), &dummy)
 			responseText += dummy.Content
 			var choice dto.ChatCompletionsStreamResponseChoice
-			choice.Delta.Content = dummy.Content
+			choice.Delta.SetContentString(dummy.Content)
 			response := dto.ChatCompletionsStreamResponse{
 				Id:      fmt.Sprintf("chatcmpl-%s", common.GetUUID()),
 				Object:  "chat.completion.chunk",
@@ -257,7 +256,7 @@ func geminiChatHandler(c *gin.Context, resp *http.Response, promptTokens int, mo
 		}, nil
 	}
 	fullTextResponse := responseGeminiChat2OpenAI(&geminiResponse)
-	completionTokens, _, _ := service.CountTokenText(geminiResponse.GetResponseText(), model, constant.ShouldCheckCompletionSensitive())
+	completionTokens, _, _ := service.CountTokenText(geminiResponse.GetResponseText(), model, false)
 	usage := dto.Usage{
 		PromptTokens:     promptTokens,
 		CompletionTokens: completionTokens,

relay/channel/lingyiwanwu/constrants.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

@@ -2,7 +2,6 @@ package ollama
 
 import (
 	"errors"
-	"fmt"
 	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
@@ -10,6 +9,7 @@ import (
 	"one-api/relay/channel"
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
+	relayconstant "one-api/relay/constant"
 	"one-api/service"
 )
 
@@ -20,7 +20,12 @@ func (a *Adaptor) Init(info *relaycommon.RelayInfo, request dto.GeneralOpenAIReq
 }
 
 func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
-	return fmt.Sprintf("%s/api/chat", info.BaseUrl), nil
+	switch info.RelayMode {
+	case relayconstant.RelayModeEmbeddings:
+		return info.BaseUrl + "/api/embeddings", nil
+	default:
+		return relaycommon.GetFullRequestURL(info.BaseUrl, info.RequestURLPath, info.ChannelType), nil
+	}
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, info *relaycommon.RelayInfo) error {
@@ -32,20 +37,29 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *dto.Gen
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	return requestOpenAI2Ollama(*request), nil
+	switch relayMode {
+	case relayconstant.RelayModeEmbeddings:
+		return requestOpenAI2Embeddings(*request), nil
+	default:
+		return requestOpenAI2Ollama(*request), nil
+	}
 }
 
 func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (*http.Response, error) {
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
-		err, responseText = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
+		err, responseText, _ = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
 		usage, _ = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	} else {
-		err, usage, sensitiveResp = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		if info.RelayMode == relayconstant.RelayModeEmbeddings {
+			err, usage = ollamaEmbeddingHandler(c, resp, info.PromptTokens, info.UpstreamModelName, info.RelayMode)
+		} else {
+			err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		}
 	}
 	return
 }

relay/channel/ollama/constants.go

@@ -1,5 +1,7 @@
 package ollama
 
-var ModelList []string
+var ModelList = []string{
+	"llama3-7b",
+}
 
 var ChannelName = "ollama"

relay/channel/ollama/dto.go

@@ -3,16 +3,24 @@ package ollama
 import "one-api/dto"
 
 type OllamaRequest struct {
-	Model    string         `json:"model,omitempty"`
-	Messages []dto.Message  `json:"messages,omitempty"`
-	Stream   bool           `json:"stream,omitempty"`
-	Options  *OllamaOptions `json:"options,omitempty"`
+	Model       string        `json:"model,omitempty"`
+	Messages    []dto.Message `json:"messages,omitempty"`
+	Stream      bool          `json:"stream,omitempty"`
+	Temperature float64       `json:"temperature,omitempty"`
+	Seed        float64       `json:"seed,omitempty"`
+	Topp        float64       `json:"top_p,omitempty"`
+	TopK        int           `json:"top_k,omitempty"`
+	Stop        any           `json:"stop,omitempty"`
 }
 
-type OllamaOptions struct {
-	Temperature float64 `json:"temperature,omitempty"`
-	Seed        float64 `json:"seed,omitempty"`
-	Topp        float64 `json:"top_p,omitempty"`
-	TopK        int     `json:"top_k,omitempty"`
-	Stop        any     `json:"stop,omitempty"`
+type OllamaEmbeddingRequest struct {
+	Model  string `json:"model,omitempty"`
+	Prompt any    `json:"prompt,omitempty"`
 }
+
+type OllamaEmbeddingResponse struct {
+	Embedding []float64 `json:"embedding,omitempty"`
+}
+
+//type OllamaOptions struct {
+//}

relay/channel/ollama/relay-ollama.go

@@ -1,6 +1,16 @@
 package ollama
 
-import "one-api/dto"
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"io"
+	"net/http"
+	"one-api/dto"
+	"one-api/service"
+	"strings"
+)
 
 func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) *OllamaRequest {
 	messages := make([]dto.Message, 0, len(request.Messages))
@@ -18,15 +28,82 @@ func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) *OllamaRequest {
 		Stop, _ = request.Stop.([]string)
 	}
 	return &OllamaRequest{
-		Model:    request.Model,
-		Messages: messages,
-		Stream:   request.Stream,
-		Options: &OllamaOptions{
-			Temperature: request.Temperature,
-			Seed:        request.Seed,
-			Topp:        request.TopP,
-			TopK:        request.TopK,
-			Stop:        Stop,
-		},
+		Model:       request.Model,
+		Messages:    messages,
+		Stream:      request.Stream,
+		Temperature: request.Temperature,
+		Seed:        request.Seed,
+		Topp:        request.TopP,
+		TopK:        request.TopK,
+		Stop:        Stop,
 	}
 }
+
+func requestOpenAI2Embeddings(request dto.GeneralOpenAIRequest) *OllamaEmbeddingRequest {
+	return &OllamaEmbeddingRequest{
+		Model:  request.Model,
+		Prompt: strings.Join(request.ParseInput(), " "),
+	}
+}
+
+func ollamaEmbeddingHandler(c *gin.Context, resp *http.Response, promptTokens int, model string, relayMode int) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	var ollamaEmbeddingResponse OllamaEmbeddingResponse
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = json.Unmarshal(responseBody, &ollamaEmbeddingResponse)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	data := make([]dto.OpenAIEmbeddingResponseItem, 0, 1)
+	data = append(data, dto.OpenAIEmbeddingResponseItem{
+		Embedding: ollamaEmbeddingResponse.Embedding,
+		Object:    "embedding",
+	})
+	usage := &dto.Usage{
+		TotalTokens:      promptTokens,
+		CompletionTokens: 0,
+		PromptTokens:     promptTokens,
+	}
+	embeddingResponse := &dto.OpenAIEmbeddingResponse{
+		Object: "list",
+		Data:   data,
+		Model:  model,
+		Usage:  *usage,
+	}
+	doResponseBody, err := json.Marshal(embeddingResponse)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	resp.Body = io.NopCloser(bytes.NewBuffer(doResponseBody))
+	// We shouldn't set the header before we parse the response body, because the parse part may fail.
+	// And then we will have to send an error response, but in this case, the header has already been set.
+	// So the httpClient will be confused by the response.
+	// For example, Postman will report error, and we cannot check the response at all.
+	// Copy headers
+	for k, v := range resp.Header {
+		// 删除任何现有的相同头部，以防止重复添加头部
+		c.Writer.Header().Del(k)
+		for _, vv := range v {
+			c.Writer.Header().Add(k, vv)
+		}
+	}
+	// reset content length
+	c.Writer.Header().Del("Content-Length")
+	c.Writer.Header().Set("Content-Length", fmt.Sprintf("%d", len(doResponseBody)))
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = io.Copy(c.Writer, resp.Body)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	return nil, usage
+}

relay/channel/openai/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/dto"
 	"one-api/relay/channel"
 	"one-api/relay/channel/ai360"
+	"one-api/relay/channel/lingyiwanwu"
 	"one-api/relay/channel/moonshot"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
@@ -33,9 +34,6 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		model_ := info.UpstreamModelName
 		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")
 
 		requestURL = fmt.Sprintf("/openai/deployments/%s/%s", model_, task)
 		return relaycommon.GetFullRequestURL(info.BaseUrl, requestURL, info.ChannelType), nil
@@ -71,13 +69,15 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
-		err, responseText = OpenaiStreamHandler(c, resp, info.RelayMode)
+		var toolCount int
+		err, responseText, toolCount = OpenaiStreamHandler(c, resp, info.RelayMode)
 		usage, _ = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
+		usage.CompletionTokens += toolCount * 7
 	} else {
-		err, usage, sensitiveResp = OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		err, usage = OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
 	}
 	return
 }
@@ -88,6 +88,8 @@ func (a *Adaptor) GetModelList() []string {
 		return ai360.ModelList
 	case common.ChannelTypeMoonshot:
 		return moonshot.ModelList
+	case common.ChannelTypeLingYiWanWu:
+		return lingyiwanwu.ModelList
 	default:
 		return ModelList
 	}

relay/channel/openai/constant.go

@@ -6,7 +6,7 @@ var ModelList = []string{
 	"gpt-3.5-turbo-instruct",
 	"gpt-4", "gpt-4-0314", "gpt-4-0613", "gpt-4-1106-preview", "gpt-4-0125-preview",
 	"gpt-4-32k", "gpt-4-32k-0314", "gpt-4-32k-0613",
-	"gpt-4-turbo-preview",
+	"gpt-4-turbo-preview", "gpt-4-turbo", "gpt-4-turbo-2024-04-09",
 	"gpt-4-vision-preview",
 	"text-embedding-ada-002", "text-embedding-3-small", "text-embedding-3-large",
 	"text-curie-001", "text-babbage-001", "text-ada-001", "text-davinci-002", "text-davinci-003",

relay/channel/openai/relay-openai.go

@@ -4,14 +4,10 @@ import (
 	"bufio"
 	"bytes"
 	"encoding/json"
-	"errors"
-	"fmt"
 	"github.com/gin-gonic/gin"
 	"io"
-	"log"
 	"net/http"
 	"one-api/common"
-	"one-api/constant"
 	"one-api/dto"
 	relayconstant "one-api/relay/constant"
 	"one-api/service"
@@ -20,9 +16,10 @@ import (
 	"time"
 )
 
-func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*dto.OpenAIErrorWithStatusCode, string) {
-	checkSensitive := constant.ShouldCheckCompletionSensitive()
+func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*dto.OpenAIErrorWithStatusCode, string, int) {
+	//checkSensitive := constant.ShouldCheckCompletionSensitive()
 	var responseTextBuilder strings.Builder
+	toolCount := 0
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
@@ -53,20 +50,11 @@ func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*d
 			if data[:6] != "data: " && data[:6] != "[DONE]" {
 				continue
 			}
-			sensitive := false
-			if checkSensitive {
-				// check sensitive
-				sensitive, _, data = service.SensitiveWordReplace(data, false)
-			}
-			dataChan <- data
+			common.SafeSendString(dataChan, data)
 			data = data[6:]
 			if !strings.HasPrefix(data, "[DONE]") {
 				streamItems = append(streamItems, data)
 			}
-			if sensitive && constant.StopOnSensitiveEnabled {
-				dataChan <- "data: [DONE]"
-				break
-			}
 		}
 		streamResp := "[" + strings.Join(streamItems, ",") + "]"
 		switch relayMode {
@@ -75,11 +63,38 @@ func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*d
 			err := json.Unmarshal(common.StringToByteSlice(streamResp), &streamResponses)
 			if err != nil {
 				common.SysError("error unmarshalling stream response: " + err.Error())
-				return // just ignore the error
-			}
-			for _, streamResponse := range streamResponses {
-				for _, choice := range streamResponse.Choices {
-					responseTextBuilder.WriteString(choice.Delta.Content)
+				for _, item := range streamItems {
+					var streamResponse dto.ChatCompletionsStreamResponseSimple
+					err := json.Unmarshal(common.StringToByteSlice(item), &streamResponse)
+					if err == nil {
+						for _, choice := range streamResponse.Choices {
+							responseTextBuilder.WriteString(choice.Delta.GetContentString())
+							if choice.Delta.ToolCalls != nil {
+								if len(choice.Delta.ToolCalls) > toolCount {
+									toolCount = len(choice.Delta.ToolCalls)
+								}
+								for _, tool := range choice.Delta.ToolCalls {
+									responseTextBuilder.WriteString(tool.Function.Name)
+									responseTextBuilder.WriteString(tool.Function.Arguments)
+								}
+							}
+						}
+					}
+				}
+			} else {
+				for _, streamResponse := range streamResponses {
+					for _, choice := range streamResponse.Choices {
+						responseTextBuilder.WriteString(choice.Delta.GetContentString())
+						if choice.Delta.ToolCalls != nil {
+							if len(choice.Delta.ToolCalls) > toolCount {
+								toolCount = len(choice.Delta.ToolCalls)
+							}
+							for _, tool := range choice.Delta.ToolCalls {
+								responseTextBuilder.WriteString(tool.Function.Name)
+								responseTextBuilder.WriteString(tool.Function.Arguments)
+							}
+						}
+					}
 				}
 			}
 		case relayconstant.RelayModeCompletions:
@@ -87,11 +102,20 @@ func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*d
 			err := json.Unmarshal(common.StringToByteSlice(streamResp), &streamResponses)
 			if err != nil {
 				common.SysError("error unmarshalling stream response: " + err.Error())
-				return // just ignore the error
-			}
-			for _, streamResponse := range streamResponses {
-				for _, choice := range streamResponse.Choices {
-					responseTextBuilder.WriteString(choice.Text)
+				for _, item := range streamItems {
+					var streamResponse dto.CompletionsStreamResponse
+					err := json.Unmarshal(common.StringToByteSlice(item), &streamResponse)
+					if err == nil {
+						for _, choice := range streamResponse.Choices {
+							responseTextBuilder.WriteString(choice.Text)
+						}
+					}
+				}
+			} else {
+				for _, streamResponse := range streamResponses {
+					for _, choice := range streamResponse.Choices {
+						responseTextBuilder.WriteString(choice.Text)
+					}
 				}
 			}
 		}
@@ -99,7 +123,7 @@ func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*d
 			// wait data out
 			time.Sleep(2 * time.Second)
 		}
-		common.SafeSend(stopChan, true)
+		common.SafeSendBool(stopChan, true)
 	}()
 	service.SetEventStreamHeaders(c)
 	c.Stream(func(w io.Writer) bool {
@@ -118,91 +142,62 @@ func OpenaiStreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*d
 	})
 	err := resp.Body.Close()
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), "", toolCount
 	}
 	wg.Wait()
-	return nil, responseTextBuilder.String()
+	return nil, responseTextBuilder.String(), toolCount
 }
 
-func OpenaiHandler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*dto.OpenAIErrorWithStatusCode, *dto.Usage, *dto.SensitiveResponse) {
-	var textResponse dto.TextResponse
+func OpenaiHandler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*dto.OpenAIErrorWithStatusCode, *dto.Usage) {
+	var simpleResponse dto.SimpleResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil, nil
+		return service.OpenAIErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
 	}
 	err = resp.Body.Close()
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil, nil
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
-	err = json.Unmarshal(responseBody, &textResponse)
+	err = json.Unmarshal(responseBody, &simpleResponse)
 	if err != nil {
-		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil, nil
+		return service.OpenAIErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
-	log.Printf("textResponse: %+v", textResponse)
-	if textResponse.Error != nil {
+	if simpleResponse.Error.Type != "" {
 		return &dto.OpenAIErrorWithStatusCode{
-			Error:      *textResponse.Error,
+			Error:      simpleResponse.Error,
 			StatusCode: resp.StatusCode,
-		}, nil, nil
+		}, nil
+	}
+	// Reset response body
+	resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
+	// We shouldn't set the header before we parse the response body, because the parse part may fail.
+	// And then we will have to send an error response, but in this case, the header has already been set.
+	// So the httpClient will be confused by the response.
+	// For example, Postman will report error, and we cannot check the response at all.
+	for k, v := range resp.Header {
+		c.Writer.Header().Set(k, v[0])
+	}
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = io.Copy(c.Writer, resp.Body)
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
 
-	checkSensitive := constant.ShouldCheckCompletionSensitive()
-	sensitiveWords := make([]string, 0)
-	triggerSensitive := false
-
-	if textResponse.Usage.TotalTokens == 0 || checkSensitive {
+	if simpleResponse.Usage.TotalTokens == 0 {
 		completionTokens := 0
-		for _, choice := range textResponse.Choices {
-			stringContent := string(choice.Message.Content)
-			ctkm, _, _ := service.CountTokenText(stringContent, model, false)
+		for _, choice := range simpleResponse.Choices {
+			ctkm, _, _ := service.CountTokenText(string(choice.Message.Content), model, false)
 			completionTokens += ctkm
-			if checkSensitive {
-				sensitive, words, stringContent := service.SensitiveWordReplace(stringContent, false)
-				if sensitive {
-					triggerSensitive = true
-					msg := choice.Message
-					msg.Content = common.StringToByteSlice(stringContent)
-					choice.Message = msg
-					sensitiveWords = append(sensitiveWords, words...)
-				}
-			}
 		}
-		textResponse.Usage = dto.Usage{
+		simpleResponse.Usage = dto.Usage{
 			PromptTokens:     promptTokens,
 			CompletionTokens: completionTokens,
 			TotalTokens:      promptTokens + completionTokens,
 		}
 	}
-
-	if constant.StopOnSensitiveEnabled {
-
-	} else {
-		responseBody, err = json.Marshal(textResponse)
-		// Reset response body
-		resp.Body = io.NopCloser(bytes.NewBuffer(responseBody))
-		// We shouldn't set the header before we parse the response body, because the parse part may fail.
-		// And then we will have to send an error response, but in this case, the header has already been set.
-		// So the httpClient will be confused by the response.
-		// For example, Postman will report error, and we cannot check the response at all.
-		for k, v := range resp.Header {
-			c.Writer.Header().Set(k, v[0])
-		}
-		c.Writer.WriteHeader(resp.StatusCode)
-		_, err = io.Copy(c.Writer, resp.Body)
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "copy_response_body_failed", http.StatusInternalServerError), nil, nil
-		}
-		err = resp.Body.Close()
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil, nil
-		}
-	}
-
-	if checkSensitive && triggerSensitive {
-		sensitiveWords = common.RemoveDuplicate(sensitiveWords)
-		return service.OpenAIErrorWrapper(errors.New(fmt.Sprintf("sensitive words detected: %s", strings.Join(sensitiveWords, ", "))), "sensitive_words_detected", http.StatusBadRequest), &textResponse.Usage, &dto.SensitiveResponse{
-			SensitiveWords: sensitiveWords,
-		}
-	}
-	return nil, &textResponse.Usage, nil
+	return nil, &simpleResponse.Usage
 }

relay/channel/palm/adaptor.go

@@ -39,7 +39,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
 		err, responseText = palmStreamHandler(c, resp)

relay/channel/palm/relay-palm.go

@@ -7,7 +7,6 @@ import (
 	"io"
 	"net/http"
 	"one-api/common"
-	"one-api/constant"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	"one-api/service"
@@ -62,7 +61,7 @@ func responsePaLM2OpenAI(response *PaLMChatResponse) *dto.OpenAITextResponse {
 func streamResponsePaLM2OpenAI(palmResponse *PaLMChatResponse) *dto.ChatCompletionsStreamResponse {
 	var choice dto.ChatCompletionsStreamResponseChoice
 	if len(palmResponse.Candidates) > 0 {
-		choice.Delta.Content = palmResponse.Candidates[0].Content
+		choice.Delta.SetContentString(palmResponse.Candidates[0].Content)
 	}
 	choice.FinishReason = &relaycommon.StopFinishReason
 	var response dto.ChatCompletionsStreamResponse
@@ -157,7 +156,7 @@ func palmHandler(c *gin.Context, resp *http.Response, promptTokens int, model st
 		}, nil
 	}
 	fullTextResponse := responsePaLM2OpenAI(&palmResponse)
-	completionTokens, _, _ := service.CountTokenText(palmResponse.Candidates[0].Content, model, constant.ShouldCheckCompletionSensitive())
+	completionTokens, _, _ := service.CountTokenText(palmResponse.Candidates[0].Content, model, false)
 	usage := dto.Usage{
 		PromptTokens:     promptTokens,
 		CompletionTokens: completionTokens,

relay/channel/perplexity/adaptor.go

@@ -43,13 +43,13 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
-		err, responseText = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
+		err, responseText, _ = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
 		usage, _ = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
 	} else {
-		err, usage, sensitiveResp = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
 	}
 	return
 }

relay/channel/tencent/adaptor.go

@@ -53,7 +53,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
 		err, responseText = tencentStreamHandler(c, resp)

relay/channel/tencent/relay-tencent.go

@@ -86,7 +86,7 @@ func streamResponseTencent2OpenAI(TencentResponse *TencentChatResponse) *dto.Cha
 	}
 	if len(TencentResponse.Choices) > 0 {
 		var choice dto.ChatCompletionsStreamResponseChoice
-		choice.Delta.Content = TencentResponse.Choices[0].Delta.Content
+		choice.Delta.SetContentString(TencentResponse.Choices[0].Delta.Content)
 		if TencentResponse.Choices[0].FinishReason == "stop" {
 			choice.FinishReason = &relaycommon.StopFinishReason
 		}
@@ -138,7 +138,7 @@ func tencentStreamHandler(c *gin.Context, resp *http.Response) (*dto.OpenAIError
 			}
 			response := streamResponseTencent2OpenAI(&TencentResponse)
 			if len(response.Choices) != 0 {
-				responseText += response.Choices[0].Delta.Content
+				responseText += response.Choices[0].Delta.GetContentString()
 			}
 			jsonResponse, err := json.Marshal(response)
 			if err != nil {

relay/channel/xunfei/adaptor.go

@@ -43,13 +43,13 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return dummyResp, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	splits := strings.Split(info.ApiKey, "|")
 	if len(splits) != 3 {
-		return nil, service.OpenAIErrorWrapper(errors.New("invalid auth"), "invalid_auth", http.StatusBadRequest), nil
+		return nil, service.OpenAIErrorWrapper(errors.New("invalid auth"), "invalid_auth", http.StatusBadRequest)
 	}
 	if a.request == nil {
-		return nil, service.OpenAIErrorWrapper(errors.New("request is nil"), "request_is_nil", http.StatusBadRequest), nil
+		return nil, service.OpenAIErrorWrapper(errors.New("request is nil"), "request_is_nil", http.StatusBadRequest)
 	}
 	if info.IsStream {
 		err, usage = xunfeiStreamHandler(c, *a.request, splits[0], splits[1], splits[2])

relay/channel/xunfei/relay-xunfei.go

@@ -87,7 +87,7 @@ func streamResponseXunfei2OpenAI(xunfeiResponse *XunfeiChatResponse) *dto.ChatCo
 		}
 	}
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = xunfeiResponse.Payload.Choices.Text[0].Content
+	choice.Delta.SetContentString(xunfeiResponse.Payload.Choices.Text[0].Content)
 	if xunfeiResponse.Payload.Choices.Status == 2 {
 		choice.FinishReason = &relaycommon.StopFinishReason
 	}
@@ -179,7 +179,13 @@ func xunfeiHandler(c *gin.Context, textRequest dto.GeneralOpenAIRequest, appId s
 		case stop = <-stopChan:
 		}
 	}
-
+	if len(xunfeiResponse.Payload.Choices.Text) == 0 {
+		xunfeiResponse.Payload.Choices.Text = []XunfeiChatResponseTextItem{
+			{
+				Content: "",
+			},
+		}
+	}
 	xunfeiResponse.Payload.Choices.Text[0].Content = content
 
 	response := responseXunfei2OpenAI(&xunfeiResponse)

relay/channel/zhipu/adaptor.go

@@ -46,7 +46,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		err, usage = zhipuStreamHandler(c, resp)
 	} else {

relay/channel/zhipu/relay-zhipu.go

@@ -126,7 +126,7 @@ func responseZhipu2OpenAI(response *ZhipuResponse) *dto.OpenAITextResponse {
 
 func streamResponseZhipu2OpenAI(zhipuResponse string) *dto.ChatCompletionsStreamResponse {
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = zhipuResponse
+	choice.Delta.SetContentString(zhipuResponse)
 	response := dto.ChatCompletionsStreamResponse{
 		Object:  "chat.completion.chunk",
 		Created: common.GetTimestamp(),
@@ -138,7 +138,7 @@ func streamResponseZhipu2OpenAI(zhipuResponse string) *dto.ChatCompletionsStream
 
 func streamMetaResponseZhipu2OpenAI(zhipuResponse *ZhipuStreamMetaResponse) (*dto.ChatCompletionsStreamResponse, *dto.Usage) {
 	var choice dto.ChatCompletionsStreamResponseChoice
-	choice.Delta.Content = ""
+	choice.Delta.SetContentString("")
 	choice.FinishReason = &relaycommon.StopFinishReason
 	response := dto.ChatCompletionsStreamResponse{
 		Id:      zhipuResponse.RequestId,

relay/channel/zhipu_4v/adaptor.go

@@ -44,13 +44,15 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode, sensitiveResp *dto.SensitiveResponse) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage *dto.Usage, err *dto.OpenAIErrorWithStatusCode) {
 	if info.IsStream {
 		var responseText string
-		err, responseText = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
+		var toolCount int
+		err, responseText, toolCount = openai.OpenaiStreamHandler(c, resp, info.RelayMode)
 		usage, _ = service.ResponseText2Usage(responseText, info.UpstreamModelName, info.PromptTokens)
+		usage.CompletionTokens += toolCount * 7
 	} else {
-		err, usage, sensitiveResp = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
 	}
 	return
 }

relay/channel/zhipu_4v/relay-zhipu_v4.go

@@ -74,6 +74,25 @@ func getZhipuToken(apikey string) string {
 func requestOpenAI2Zhipu(request dto.GeneralOpenAIRequest) *dto.GeneralOpenAIRequest {
 	messages := make([]dto.Message, 0, len(request.Messages))
 	for _, message := range request.Messages {
+		if !message.IsStringContent() {
+			mediaMessages := message.ParseContent()
+			for j, mediaMessage := range mediaMessages {
+				if mediaMessage.Type == dto.ContentTypeImageURL {
+					imageUrl := mediaMessage.ImageUrl.(dto.MessageImageUrl)
+					// check if base64
+					if strings.HasPrefix(imageUrl.Url, "data:image/") {
+						// 去除base64数据的URL前缀（如果有）
+						if idx := strings.Index(imageUrl.Url, ","); idx != -1 {
+							imageUrl.Url = imageUrl.Url[idx+1:]
+						}
+					}
+					mediaMessage.ImageUrl = imageUrl
+					mediaMessages[j] = mediaMessage
+				}
+			}
+			messageRaw, _ := json.Marshal(mediaMessages)
+			message.Content = messageRaw
+		}
 		messages = append(messages, dto.Message{
 			Role:       message.Role,
 			Content:    message.Content,
@@ -138,7 +157,7 @@ func streamResponseZhipu2OpenAI(zhipuResponse *ZhipuV4StreamResponse) *dto.ChatC
 		Id:      zhipuResponse.Id,
 		Object:  "chat.completion.chunk",
 		Created: zhipuResponse.Created,
-		Model:   "glm-4",
+		Model:   "glm-4v",
 		Choices: []dto.ChatCompletionsStreamResponseChoice{choice},
 	}
 	return &response

relay/common/relay_info.go

@@ -31,6 +31,7 @@ type RelayInfo struct {
 func GenRelayInfo(c *gin.Context) *RelayInfo {
 	channelType := c.GetInt("channel")
 	channelId := c.GetInt("channel_id")
+
 	tokenId := c.GetInt("token_id")
 	userId := c.GetInt("id")
 	group := c.GetString("group")

relay/common/relay_utils.go

@@ -35,12 +35,12 @@ func RelayErrorHandler(resp *http.Response) (OpenAIErrorWithStatusCode *dto.Open
 	if err != nil {
 		return
 	}
-	var textResponse dto.TextResponse
+	var textResponse dto.TextResponseWithError
 	err = json.Unmarshal(responseBody, &textResponse)
 	if err != nil {
 		return
 	}
-	OpenAIErrorWithStatusCode.Error = *textResponse.Error
+	OpenAIErrorWithStatusCode.Error = textResponse.Error
 	return
 }
 

relay/constant/api_type.go

@@ -18,13 +18,25 @@ const (
 	APITypeZhipu_v4
 	APITypeOllama
 	APITypePerplexity
+	APITypeAws
+	APITypeCohere
 
 	APITypeDummy // this one is only for count, do not add any channel after this
 )
 
 func ChannelType2APIType(channelType int) int {
-	apiType := APITypeOpenAI
+	apiType := -1
 	switch channelType {
+	case common.ChannelTypeOpenAI:
+		apiType = APITypeOpenAI
+	case common.ChannelTypeAzure:
+		apiType = APITypeOpenAI
+	case common.ChannelTypeMoonshot:
+		apiType = APITypeOpenAI
+	case common.ChannelTypeLingYiWanWu:
+		apiType = APITypeOpenAI
+	case common.ChannelType360:
+		apiType = APITypeOpenAI
 	case common.ChannelTypeAnthropic:
 		apiType = APITypeAnthropic
 	case common.ChannelTypeBaidu:
@@ -49,6 +61,10 @@ func ChannelType2APIType(channelType int) int {
 		apiType = APITypeOllama
 	case common.ChannelTypePerplexity:
 		apiType = APITypePerplexity
+	case common.ChannelTypeAws:
+		apiType = APITypeAws
+	case common.ChannelTypeCohere:
+		apiType = APITypeCohere
 	}
 	return apiType
 }

relay/constant/relay_mode.go

@@ -56,29 +56,29 @@ func Path2RelayMode(path string) int {
 
 func Path2RelayModeMidjourney(path string) int {
 	relayMode := RelayModeUnknown
-	if strings.HasPrefix(path, "/mj/submit/action") {
+	if strings.HasSuffix(path, "/mj/submit/action") {
 		// midjourney plus
 		relayMode = RelayModeMidjourneyAction
-	} else if strings.HasPrefix(path, "/mj/submit/modal") {
+	} else if strings.HasSuffix(path, "/mj/submit/modal") {
 		// midjourney plus
 		relayMode = RelayModeMidjourneyModal
-	} else if strings.HasPrefix(path, "/mj/submit/shorten") {
+	} else if strings.HasSuffix(path, "/mj/submit/shorten") {
 		// midjourney plus
 		relayMode = RelayModeMidjourneyShorten
-	} else if strings.HasPrefix(path, "/mj/insight-face/swap") {
+	} else if strings.HasSuffix(path, "/mj/insight-face/swap") {
 		// midjourney plus
 		relayMode = RelayModeSwapFace
-	} else if strings.HasPrefix(path, "/mj/submit/imagine") {
+	} else if strings.HasSuffix(path, "/mj/submit/imagine") {
 		relayMode = RelayModeMidjourneyImagine
-	} else if strings.HasPrefix(path, "/mj/submit/blend") {
+	} else if strings.HasSuffix(path, "/mj/submit/blend") {
 		relayMode = RelayModeMidjourneyBlend
-	} else if strings.HasPrefix(path, "/mj/submit/describe") {
+	} else if strings.HasSuffix(path, "/mj/submit/describe") {
 		relayMode = RelayModeMidjourneyDescribe
-	} else if strings.HasPrefix(path, "/mj/notify") {
+	} else if strings.HasSuffix(path, "/mj/notify") {
 		relayMode = RelayModeMidjourneyNotify
-	} else if strings.HasPrefix(path, "/mj/submit/change") {
+	} else if strings.HasSuffix(path, "/mj/submit/change") {
 		relayMode = RelayModeMidjourneyChange
-	} else if strings.HasPrefix(path, "/mj/submit/simple-change") {
+	} else if strings.HasSuffix(path, "/mj/submit/simple-change") {
 		relayMode = RelayModeMidjourneyChange
 	} else if strings.HasSuffix(path, "/fetch") {
 		relayMode = RelayModeMidjourneyTaskFetch

relay/relay-audio.go

@@ -20,15 +20,6 @@ import (
 	"time"
 )
 
-var availableVoices = []string{
-	"alloy",
-	"echo",
-	"fable",
-	"onyx",
-	"nova",
-	"shimmer",
-}
-
 func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 	tokenId := c.GetInt("token_id")
 	channelType := c.GetInt("channel")
@@ -59,9 +50,6 @@ func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 		if audioRequest.Voice == "" {
 			return service.OpenAIErrorWrapper(errors.New("voice is required"), "required_field_missing", http.StatusBadRequest)
 		}
-		if !common.StringsContains(availableVoices, audioRequest.Voice) {
-			return service.OpenAIErrorWrapper(errors.New("voice must be one of "+strings.Join(availableVoices, ", ")), "invalid_field_value", http.StatusBadRequest)
-		}
 	}
 	var err error
 	promptTokens := 0
@@ -100,6 +88,22 @@ func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 		}
 	}
 
+	succeed := false
+	defer func() {
+		if succeed {
+			return
+		}
+		if preConsumedQuota > 0 {
+			// we need to roll back the pre-consumed quota
+			defer func() {
+				go func() {
+					// negative means add quota back for token & user
+					returnPreConsumedQuota(c, tokenId, userQuota, preConsumedQuota)
+				}()
+			}()
+		}
+	}()
+
 	// map model name
 	modelMapping := c.GetString("model_mapping")
 	if modelMapping != "" {
@@ -163,6 +167,7 @@ func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 	if resp.StatusCode != http.StatusOK {
 		return relaycommon.RelayErrorHandler(resp)
 	}
+	succeed = true
 
 	var audioResponse dto.AudioResponse
 
@@ -173,7 +178,7 @@ func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 			if strings.HasPrefix(audioRequest.Model, "tts-1") {
 				quota = promptTokens
 			} else {
-				quota, err, _ = service.CountAudioToken(audioResponse.Text, audioRequest.Model, constant.ShouldCheckCompletionSensitive())
+				quota, err, _ = service.CountAudioToken(audioResponse.Text, audioRequest.Model, false)
 			}
 			quota = int(float64(quota) * ratio)
 			if ratio != 0 && quota <= 0 {
@@ -191,7 +196,10 @@ func AudioHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
 				logContent := fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f", modelRatio, groupRatio)
-				model.RecordConsumeLog(ctx, userId, channelId, promptTokens, 0, audioRequest.Model, tokenName, quota, logContent, tokenId, userQuota, int(useTimeSeconds), false)
+				other := make(map[string]interface{})
+				other["model_ratio"] = modelRatio
+				other["group_ratio"] = groupRatio
+				model.RecordConsumeLog(ctx, userId, channelId, promptTokens, 0, audioRequest.Model, tokenName, quota, logContent, tokenId, userQuota, int(useTimeSeconds), false, other)
 				model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
 				channelId := c.GetInt("channel_id")
 				model.UpdateChannelUsedQuota(channelId, quota)

relay/relay-image.go

@@ -34,7 +34,7 @@ func RelayImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusC
 	}
 
 	if imageRequest.Model == "" {
-		imageRequest.Model = "dall-e-2"
+		imageRequest.Model = "dall-e-3"
 	}
 	if imageRequest.Size == "" {
 		imageRequest.Size = "1024x1024"
@@ -186,8 +186,15 @@ func RelayImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusC
 		}
 		if quota != 0 {
 			tokenName := c.GetString("token_name")
-			logContent := fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f", modelRatio, groupRatio)
-			model.RecordConsumeLog(ctx, userId, channelId, 0, 0, imageRequest.Model, tokenName, quota, logContent, tokenId, userQuota, int(useTimeSeconds), false)
+			quality := "normal"
+			if imageRequest.Quality == "hd" {
+				quality = "hd"
+			}
+			logContent := fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f, 大小 %s, 品质 %s", modelRatio, groupRatio, imageRequest.Size, quality)
+			other := make(map[string]interface{})
+			other["model_ratio"] = modelRatio
+			other["group_ratio"] = groupRatio
+			model.RecordConsumeLog(ctx, userId, channelId, 0, 0, imageRequest.Model, tokenName, quota, logContent, tokenId, userQuota, int(useTimeSeconds), false, other)
 			model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
 			channelId := c.GetInt("channel_id")
 			model.UpdateChannelUsedQuota(channelId, quota)

relay/relay-mj.go

@@ -110,11 +110,13 @@ func coverMidjourneyTaskDto(c *gin.Context, originTask *model.Midjourney) (midjo
 	midjourneyTask.StartTime = originTask.StartTime
 	midjourneyTask.FinishTime = originTask.FinishTime
 	midjourneyTask.ImageUrl = ""
-	if originTask.ImageUrl != "" {
+	if originTask.ImageUrl != "" && constant.MjForwardUrlEnabled {
 		midjourneyTask.ImageUrl = common.ServerAddress + "/mj/image/" + originTask.MjId
 		if originTask.Status != "SUCCESS" {
 			midjourneyTask.ImageUrl += "?rand=" + strconv.FormatInt(time.Now().UnixNano(), 10)
 		}
+	} else {
+		midjourneyTask.ImageUrl = originTask.ImageUrl
 	}
 	midjourneyTask.Status = originTask.Status
 	midjourneyTask.FailReason = originTask.FailReason
@@ -180,7 +182,7 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 			Description: "quota_not_enough",
 		}
 	}
-	requestURL := c.Request.URL.String()
+	requestURL := getMjRequestPath(c.Request.URL.String())
 	baseURL := c.GetString("base_url")
 	fullRequestURL := fmt.Sprintf("%s%s", baseURL, requestURL)
 	mjResp, _, err := service.DoMidjourneyHttpRequest(c, time.Second*60, fullRequestURL)
@@ -200,7 +202,10 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
 				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, groupRatio, constant.MjActionSwapFace)
-				model.RecordConsumeLog(ctx, userId, channelId, 0, 0, modelName, tokenName, quota, logContent, tokenId, userQuota, 0, false)
+				other := make(map[string]interface{})
+				other["model_price"] = modelPrice
+				other["group_ratio"] = groupRatio
+				model.RecordConsumeLog(ctx, userId, channelId, 0, 0, modelName, tokenName, quota, logContent, tokenId, userQuota, 0, false, other)
 				model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
 				channelId := c.GetInt("channel_id")
 				model.UpdateChannelUsedQuota(channelId, quota)
@@ -260,7 +265,7 @@ func RelayMidjourneyTaskImageSeed(c *gin.Context) *dto.MidjourneyResponse {
 	c.Set("channel_id", originTask.ChannelId)
 	c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
 
-	requestURL := c.Request.URL.String()
+	requestURL := getMjRequestPath(c.Request.URL.String())
 	fullRequestURL := fmt.Sprintf("%s%s", channel.GetBaseURL(), requestURL)
 	midjResponseWithStatus, _, err := service.DoMidjourneyHttpRequest(c, time.Second*30, fullRequestURL)
 	if err != nil {
@@ -440,7 +445,7 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 	}
 
 	//baseURL := common.ChannelBaseURLs[channelType]
-	requestURL := c.Request.URL.String()
+	requestURL := getMjRequestPath(c.Request.URL.String())
 
 	baseURL := c.GetString("base_url")
 
@@ -496,7 +501,10 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 			if quota != 0 {
 				tokenName := c.GetString("token_name")
 				logContent := fmt.Sprintf("模型固定价格 %.2f，分组倍率 %.2f，操作 %s", modelPrice, groupRatio, midjRequest.Action)
-				model.RecordConsumeLog(ctx, userId, channelId, 0, 0, modelName, tokenName, quota, logContent, tokenId, userQuota, 0, false)
+				other := make(map[string]interface{})
+				other["model_price"] = modelPrice
+				other["group_ratio"] = groupRatio
+				model.RecordConsumeLog(ctx, userId, channelId, 0, 0, modelName, tokenName, quota, logContent, tokenId, userQuota, 0, false, other)
 				model.UpdateUserUsedQuotaAndRequestCount(userId, quota)
 				channelId := c.GetInt("channel_id")
 				model.UpdateChannelUsedQuota(channelId, quota)
@@ -605,3 +613,15 @@ type taskChangeParams struct {
 	Action string
 	Index  int
 }
+
+func getMjRequestPath(path string) string {
+	requestURL := path
+	if strings.Contains(requestURL, "/mj-") {
+		urls := strings.Split(requestURL, "/mj/")
+		if len(urls) < 2 {
+			return requestURL
+		}
+		requestURL = "/mj/" + urls[1]
+	}
+	return requestURL
+}

relay/relay-text.go

@@ -72,7 +72,7 @@ func TextHelper(c *gin.Context) *dto.OpenAIErrorWithStatusCode {
 	textRequest, err := getAndValidateTextRequest(c, relayInfo)
 	if err != nil {
 		common.LogError(c, fmt.Sprintf("getAndValidateTextRequest failed: %s", err.Error()))
-		return service.OpenAIErrorWrapper(err, "invalid_text_request", http.StatusBadRequest)
+		return service.OpenAIErrorWrapperLocal(err, "invalid_text_request", http.StatusBadRequest)
 	}
 
 	// map model name
@@ -82,7 +82,7 @@ func TextHelper(c *gin.Context) *dto.OpenAIErrorWithStatusCode {
 		modelMap := make(map[string]string)
 		err := json.Unmarshal([]byte(modelMapping), &modelMap)
 		if err != nil {
-			return service.OpenAIErrorWrapper(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
+			return service.OpenAIErrorWrapperLocal(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
 		}
 		if modelMap[textRequest.Model] != "" {
 			textRequest.Model = modelMap[textRequest.Model]
@@ -103,7 +103,7 @@ func TextHelper(c *gin.Context) *dto.OpenAIErrorWithStatusCode {
 	// count messages token error 计算promptTokens错误
 	if err != nil {
 		if sensitiveTrigger {
-			return service.OpenAIErrorWrapper(err, "sensitive_words_detected", http.StatusBadRequest)
+			return service.OpenAIErrorWrapperLocal(err, "sensitive_words_detected", http.StatusBadRequest)
 		}
 		return service.OpenAIErrorWrapper(err, "count_token_messages_failed", http.StatusInternalServerError)
 	}
@@ -154,32 +154,31 @@ func TextHelper(c *gin.Context) *dto.OpenAIErrorWithStatusCode {
 		requestBody = bytes.NewBuffer(jsonData)
 	}
 
+	statusCodeMappingStr := c.GetString("status_code_mapping")
 	resp, err := adaptor.DoRequest(c, relayInfo, requestBody)
 	if err != nil {
 		return service.OpenAIErrorWrapper(err, "do_request_failed", http.StatusInternalServerError)
 	}
-	relayInfo.IsStream = relayInfo.IsStream || strings.HasPrefix(resp.Header.Get("Content-Type"), "text/event-stream")
 
-	if resp.StatusCode != http.StatusOK {
-		returnPreConsumedQuota(c, relayInfo.TokenId, userQuota, preConsumedQuota)
-		return service.RelayErrorHandler(resp)
-	}
-
-	usage, openaiErr, sensitiveResp := adaptor.DoResponse(c, resp, relayInfo)
-	if openaiErr != nil {
-		if sensitiveResp == nil { // 如果没有敏感词检查结果
+	if resp != nil {
+		relayInfo.IsStream = relayInfo.IsStream || strings.HasPrefix(resp.Header.Get("Content-Type"), "text/event-stream")
+		if resp.StatusCode != http.StatusOK {
 			returnPreConsumedQuota(c, relayInfo.TokenId, userQuota, preConsumedQuota)
+			openaiErr := service.RelayErrorHandler(resp)
+			// reset status code 重置状态码
+			service.ResetStatusCode(openaiErr, statusCodeMappingStr)
 			return openaiErr
-		} else {
-			// 如果有敏感词检查结果，不返回预消耗配额，继续消耗配额
-			postConsumeQuota(c, relayInfo, *textRequest, usage, ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, sensitiveResp)
-			if constant.StopOnSensitiveEnabled { // 是否直接返回错误
-				return openaiErr
-			}
-			return nil
 		}
 	}
-	postConsumeQuota(c, relayInfo, *textRequest, usage, ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, nil)
+
+	usage, openaiErr := adaptor.DoResponse(c, resp, relayInfo)
+	if openaiErr != nil {
+		returnPreConsumedQuota(c, relayInfo.TokenId, userQuota, preConsumedQuota)
+		// reset status code 重置状态码
+		service.ResetStatusCode(openaiErr, statusCodeMappingStr)
+		return openaiErr
+	}
+	postConsumeQuota(c, relayInfo, *textRequest, usage, ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice)
 	return nil
 }
 
@@ -190,7 +189,7 @@ func getPromptTokens(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.Re
 	checkSensitive := constant.ShouldCheckPromptSensitive()
 	switch info.RelayMode {
 	case relayconstant.RelayModeChatCompletions:
-		promptTokens, err, sensitiveTrigger = service.CountTokenMessages(textRequest.Messages, textRequest.Model, checkSensitive)
+		promptTokens, err, sensitiveTrigger = service.CountTokenChatRequest(*textRequest, textRequest.Model, checkSensitive)
 	case relayconstant.RelayModeCompletions:
 		promptTokens, err, sensitiveTrigger = service.CountTokenInput(textRequest.Prompt, textRequest.Model, checkSensitive)
 	case relayconstant.RelayModeModerations:
@@ -209,14 +208,14 @@ func getPromptTokens(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.Re
 func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommon.RelayInfo) (int, int, *dto.OpenAIErrorWithStatusCode) {
 	userQuota, err := model.CacheGetUserQuota(relayInfo.UserId)
 	if err != nil {
-		return 0, 0, service.OpenAIErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
+		return 0, 0, service.OpenAIErrorWrapperLocal(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
 	if userQuota <= 0 || userQuota-preConsumedQuota < 0 {
-		return 0, 0, service.OpenAIErrorWrapper(errors.New("user quota is not enough"), "insufficient_user_quota", http.StatusForbidden)
+		return 0, 0, service.OpenAIErrorWrapperLocal(errors.New("user quota is not enough"), "insufficient_user_quota", http.StatusForbidden)
 	}
 	err = model.CacheDecreaseUserQuota(relayInfo.UserId, preConsumedQuota)
 	if err != nil {
-		return 0, 0, service.OpenAIErrorWrapper(err, "decrease_user_quota_failed", http.StatusInternalServerError)
+		return 0, 0, service.OpenAIErrorWrapperLocal(err, "decrease_user_quota_failed", http.StatusInternalServerError)
 	}
 	if userQuota > 100*preConsumedQuota {
 		// 用户额度充足，判断令牌额度是否充足
@@ -238,7 +237,7 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 	if preConsumedQuota > 0 {
 		userQuota, err = model.PreConsumeTokenQuota(relayInfo.TokenId, preConsumedQuota)
 		if err != nil {
-			return 0, 0, service.OpenAIErrorWrapper(err, "pre_consume_token_quota_failed", http.StatusForbidden)
+			return 0, 0, service.OpenAIErrorWrapperLocal(err, "pre_consume_token_quota_failed", http.StatusForbidden)
 		}
 	}
 	return preConsumedQuota, userQuota, nil
@@ -258,17 +257,17 @@ func returnPreConsumedQuota(c *gin.Context, tokenId int, userQuota int, preConsu
 
 func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, textRequest dto.GeneralOpenAIRequest,
 	usage *dto.Usage, ratio float64, preConsumedQuota int, userQuota int, modelRatio float64, groupRatio float64,
-	modelPrice float64, sensitiveResp *dto.SensitiveResponse) {
+	modelPrice float64) {
 
 	useTimeSeconds := time.Now().Unix() - relayInfo.StartTime.Unix()
 	promptTokens := usage.PromptTokens
 	completionTokens := usage.CompletionTokens
 
 	tokenName := ctx.GetString("token_name")
+	completionRatio := common.GetCompletionRatio(textRequest.Model)
 
 	quota := 0
 	if modelPrice == -1 {
-		completionRatio := common.GetCompletionRatio(textRequest.Model)
 		quota = promptTokens + int(float64(completionTokens)*completionRatio)
 		quota = int(float64(quota) * ratio)
 		if ratio != 0 && quota <= 0 {
@@ -280,7 +279,7 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, textRe
 	totalTokens := promptTokens + completionTokens
 	var logContent string
 	if modelPrice == -1 {
-		logContent = fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f", modelRatio, groupRatio)
+		logContent = fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f，补全倍率 %.2f", modelRatio, groupRatio, completionRatio)
 	} else {
 		logContent = fmt.Sprintf("模型价格 %.2f，分组倍率 %.2f", modelPrice, groupRatio)
 	}
@@ -293,15 +292,17 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, textRe
 		logContent += fmt.Sprintf("（可能是上游超时）")
 		common.LogError(ctx, fmt.Sprintf("total tokens is 0, cannot consume quota, userId %d, channelId %d, tokenId %d, model %s， pre-consumed quota %d", relayInfo.UserId, relayInfo.ChannelId, relayInfo.TokenId, textRequest.Model, preConsumedQuota))
 	} else {
-		if sensitiveResp != nil {
-			logContent += fmt.Sprintf("，敏感词：%s", strings.Join(sensitiveResp.SensitiveWords, ", "))
-		}
+		//if sensitiveResp != nil {
+		//	logContent += fmt.Sprintf("，敏感词：%s", strings.Join(sensitiveResp.SensitiveWords, ", "))
+		//}
 		quotaDelta := quota - preConsumedQuota
-		err := model.PostConsumeTokenQuota(relayInfo.TokenId, userQuota, quotaDelta, preConsumedQuota, true)
-		if err != nil {
-			common.LogError(ctx, "error consuming token remain quota: "+err.Error())
+		if quotaDelta != 0 {
+			err := model.PostConsumeTokenQuota(relayInfo.TokenId, userQuota, quotaDelta, preConsumedQuota, true)
+			if err != nil {
+				common.LogError(ctx, "error consuming token remain quota: "+err.Error())
+			}
 		}
-		err = model.CacheUpdateUserQuota(relayInfo.UserId)
+		err := model.CacheUpdateUserQuota(relayInfo.UserId)
 		if err != nil {
 			common.LogError(ctx, "error update user quota cache: "+err.Error())
 		}
@@ -314,7 +315,12 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, textRe
 		logModel = "gpt-4-gizmo-*"
 		logContent += fmt.Sprintf("，模型 %s", textRequest.Model)
 	}
-	model.RecordConsumeLog(ctx, relayInfo.UserId, relayInfo.ChannelId, promptTokens, completionTokens, logModel, tokenName, quota, logContent, relayInfo.TokenId, userQuota, int(useTimeSeconds), relayInfo.IsStream)
+	other := make(map[string]interface{})
+	other["model_ratio"] = modelRatio
+	other["group_ratio"] = groupRatio
+	other["completion_ratio"] = completionRatio
+	other["model_price"] = modelPrice
+	model.RecordConsumeLog(ctx, relayInfo.UserId, relayInfo.ChannelId, promptTokens, completionTokens, logModel, tokenName, quota, logContent, relayInfo.TokenId, userQuota, int(useTimeSeconds), relayInfo.IsStream, other)
 
 	//if quota != 0 {
 	//

relay/relay_adaptor.go

@@ -3,8 +3,10 @@ package relay
 import (
 	"one-api/relay/channel"
 	"one-api/relay/channel/ali"
+	"one-api/relay/channel/aws"
 	"one-api/relay/channel/baidu"
 	"one-api/relay/channel/claude"
+	"one-api/relay/channel/cohere"
 	"one-api/relay/channel/gemini"
 	"one-api/relay/channel/ollama"
 	"one-api/relay/channel/openai"
@@ -45,6 +47,10 @@ func GetAdaptor(apiType int) channel.Adaptor {
 		return &ollama.Adaptor{}
 	case constant.APITypePerplexity:
 		return &perplexity.Adaptor{}
+	case constant.APITypeAws:
+		return &aws.Adaptor{}
+	case constant.APITypeCohere:
+		return &cohere.Adaptor{}
 	}
 	return nil
 }

router/api-router.go

@@ -14,10 +14,11 @@ func SetApiRouter(router *gin.Engine) {
 	apiRouter.Use(middleware.GlobalAPIRateLimit())
 	{
 		apiRouter.GET("/status", controller.GetStatus)
+		apiRouter.GET("/models", middleware.UserAuth(), controller.DashboardListModels)
 		apiRouter.GET("/status/test", middleware.AdminAuth(), controller.TestStatus)
 		apiRouter.GET("/notice", controller.GetNotice)
 		apiRouter.GET("/about", controller.GetAbout)
-		apiRouter.GET("/midjourney", controller.GetMidjourney)
+		//apiRouter.GET("/midjourney", controller.GetMidjourney)
 		apiRouter.GET("/home_page_content", controller.GetHomePageContent)
 		apiRouter.GET("/verification", middleware.CriticalRateLimit(), middleware.TurnstileCheck(), controller.SendEmailVerification)
 		apiRouter.GET("/reset_password", middleware.CriticalRateLimit(), middleware.TurnstileCheck(), controller.SendPasswordResetEmail)

router/relay-router.go

@@ -43,7 +43,16 @@ func SetRelayRouter(router *gin.Engine) {
 		relayV1Router.DELETE("/models/:model", controller.RelayNotImplemented)
 		relayV1Router.POST("/moderations", controller.Relay)
 	}
+
 	relayMjRouter := router.Group("/mj")
+	registerMjRouterGroup(relayMjRouter)
+
+	relayMjModeRouter := router.Group("/:mode/mj")
+	registerMjRouterGroup(relayMjModeRouter)
+	//relayMjRouter.Use()
+}
+
+func registerMjRouterGroup(relayMjRouter *gin.RouterGroup) {
 	relayMjRouter.GET("/image/:id", relay.RelayMidjourneyImage)
 	relayMjRouter.Use(middleware.TokenAuth(), middleware.Distribute())
 	{
@@ -61,5 +70,4 @@ func SetRelayRouter(router *gin.Engine) {
 		relayMjRouter.POST("/task/list-by-condition", controller.RelayMidjourney)
 		relayMjRouter.POST("/insight-face/swap", controller.RelayMidjourney)
 	}
-	//relayMjRouter.Use()
 }

router/web-router.go

@@ -16,9 +16,9 @@ func SetWebRouter(router *gin.Engine, buildFS embed.FS, indexPage []byte) {
 	router.Use(gzip.Gzip(gzip.DefaultCompression))
 	router.Use(middleware.GlobalWebRateLimit())
 	router.Use(middleware.Cache())
-	router.Use(static.Serve("/", common.EmbedFolder(buildFS, "web/build")))
+	router.Use(static.Serve("/", common.EmbedFolder(buildFS, "web/dist")))
 	router.NoRoute(func(c *gin.Context) {
-		if strings.HasPrefix(c.Request.RequestURI, "/v1") || strings.HasPrefix(c.Request.RequestURI, "/api") {
+		if strings.HasPrefix(c.Request.RequestURI, "/v1") || strings.HasPrefix(c.Request.RequestURI, "/api") || strings.HasPrefix(c.Request.RequestURI, "/assets") {
 			controller.RelayNotFound(c)
 			return
 		}

service/channel.go

@@ -6,6 +6,7 @@ import (
 	"one-api/common"
 	relaymodel "one-api/dto"
 	"one-api/model"
+	"strings"
 )
 
 // disable & notify
@@ -33,13 +34,36 @@ func ShouldDisableChannel(err *relaymodel.OpenAIError, statusCode int) bool {
 	if statusCode == http.StatusUnauthorized {
 		return true
 	}
-	if err.Type == "insufficient_quota" || err.Code == "invalid_api_key" || err.Code == "account_deactivated" || err.Code == "billing_not_active" {
+	switch err.Code {
+	case "invalid_api_key":
+		return true
+	case "account_deactivated":
+		return true
+	case "billing_not_active":
+		return true
+	}
+	switch err.Type {
+	case "insufficient_quota":
+		return true
+	// https://docs.anthropic.com/claude/reference/errors
+	case "authentication_error":
+		return true
+	case "permission_error":
+		return true
+	case "forbidden":
+		return true
+	}
+	if strings.HasPrefix(err.Message, "Your credit balance is too low") { // anthropic
+		return true
+	} else if strings.HasPrefix(err.Message, "This organization has been disabled.") {
+		return true
+	} else if strings.HasPrefix(err.Message, "You exceeded your current quota") {
 		return true
 	}
 	return false
 }
 
-func ShouldEnableChannel(err error, openAIErr *relaymodel.OpenAIError) bool {
+func ShouldEnableChannel(err error, openAIErr *relaymodel.OpenAIError, status int) bool {
 	if !common.AutomaticEnableChannelEnabled {
 		return false
 	}
@@ -49,5 +73,8 @@ func ShouldEnableChannel(err error, openAIErr *relaymodel.OpenAIError) bool {
 	if openAIErr != nil {
 		return false
 	}
+	if status != common.ChannelStatusAutoDisabled {
+		return false
+	}
 	return true
 }

service/epay.go

@@ -1,10 +1,13 @@
 package service
 
-import "one-api/common"
+import (
+	"one-api/common"
+	"one-api/constant"
+)
 
 func GetCallbackAddress() string {
-	if common.CustomCallbackAddress == "" {
+	if constant.CustomCallbackAddress == "" {
 		return common.ServerAddress
 	}
-	return common.CustomCallbackAddress
+	return constant.CustomCallbackAddress
 }

service/error.go

@@ -29,7 +29,7 @@ func MidjourneyErrorWithStatusCodeWrapper(code int, desc string, statusCode int)
 func OpenAIErrorWrapper(err error, code string, statusCode int) *dto.OpenAIErrorWithStatusCode {
 	text := err.Error()
 	// 定义一个正则表达式匹配URL
-	if strings.Contains(text, "Post") {
+	if strings.Contains(text, "Post") || strings.Contains(text, "dial") {
 		common.SysLog(fmt.Sprintf("error: %s", text))
 		text = "请求上游地址失败"
 	}
@@ -46,6 +46,12 @@ func OpenAIErrorWrapper(err error, code string, statusCode int) *dto.OpenAIError
 	}
 }
 
+func OpenAIErrorWrapperLocal(err error, code string, statusCode int) *dto.OpenAIErrorWithStatusCode {
+	openaiErr := OpenAIErrorWrapper(err, code, statusCode)
+	openaiErr.LocalError = true
+	return openaiErr
+}
+
 func RelayErrorHandler(resp *http.Response) (errWithStatusCode *dto.OpenAIErrorWithStatusCode) {
 	errWithStatusCode = &dto.OpenAIErrorWithStatusCode{
 		StatusCode: resp.StatusCode,
@@ -80,3 +86,22 @@ func RelayErrorHandler(resp *http.Response) (errWithStatusCode *dto.OpenAIErrorW
 	}
 	return
 }
+
+func ResetStatusCode(openaiErr *dto.OpenAIErrorWithStatusCode, statusCodeMappingStr string) {
+	if statusCodeMappingStr == "" || statusCodeMappingStr == "{}" {
+		return
+	}
+	statusCodeMapping := make(map[string]string)
+	err := json.Unmarshal([]byte(statusCodeMappingStr), &statusCodeMapping)
+	if err != nil {
+		return
+	}
+	if openaiErr.StatusCode == http.StatusOK {
+		return
+	}
+	codeStr := strconv.Itoa(openaiErr.StatusCode)
+	if _, ok := statusCodeMapping[codeStr]; ok {
+		intCode, _ := strconv.Atoi(statusCodeMapping[codeStr])
+		openaiErr.StatusCode = intCode
+	}
+}

service/midjourney.go

@@ -165,13 +165,24 @@ func DoMidjourneyHttpRequest(c *gin.Context, timeout time.Duration, fullRequestU
 		if err != nil {
 			return MidjourneyErrorWithStatusCodeWrapper(constant.MjErrorUnknown, "read_request_body_failed", http.StatusInternalServerError), nullBytes, err
 		}
-		delete(mapResult, "accountFilter")
+		if !constant.MjAccountFilterEnabled {
+			delete(mapResult, "accountFilter")
+		}
 		if !constant.MjNotifyEnabled {
 			delete(mapResult, "notifyHook")
 		}
 		//req, err := http.NewRequest(c.Request.Method, fullRequestURL, requestBody)
 		// make new request with mapResult
 	}
+	if constant.MjModeClearEnabled {
+		if prompt, ok := mapResult["prompt"].(string); ok {
+			prompt = strings.Replace(prompt, "--fast", "", -1)
+			prompt = strings.Replace(prompt, "--relax", "", -1)
+			prompt = strings.Replace(prompt, "--turbo", "", -1)
+
+			mapResult["prompt"] = prompt
+		}
+	}
 	reqBody, err := json.Marshal(mapResult)
 	if err != nil {
 		return MidjourneyErrorWithStatusCodeWrapper(constant.MjErrorUnknown, "marshal_request_body_failed", http.StatusInternalServerError), nullBytes, err
@@ -185,7 +196,11 @@ func DoMidjourneyHttpRequest(c *gin.Context, timeout time.Duration, fullRequestU
 	req = req.WithContext(ctx)
 	req.Header.Set("Content-Type", c.Request.Header.Get("Content-Type"))
 	req.Header.Set("Accept", c.Request.Header.Get("Accept"))
-	req.Header.Set("mj-api-secret", strings.Split(c.Request.Header.Get("Authorization"), " ")[1])
+	auth := c.Request.Header.Get("Authorization")
+	if auth != "" {
+		auth = strings.TrimPrefix(auth, "Bearer ")
+		req.Header.Set("mj-api-secret", auth)
+	}
 	defer cancel()
 	resp, err := GetHttpClient().Do(req)
 	if err != nil {

service/sensitive.go

@@ -40,7 +40,7 @@ func SensitiveWordReplace(text string, returnImmediately bool) (bool, []string,
 		for _, hit := range hits {
 			pos := hit.Pos
 			word := string(hit.Word)
-			text = text[:pos] + "*###*" + text[pos+len(word):]
+			text = text[:pos] + "**###**" + text[pos+len(word):]
 			words = append(words, word)
 		}
 		return true, words, text

service/token_counter.go

@@ -29,7 +29,7 @@ func InitTokenEncoders() {
 	if err != nil {
 		common.FatalLog(fmt.Sprintf("failed to get gpt-4 token encoder: %s", err.Error()))
 	}
-	for model, _ := range common.ModelRatio {
+	for model, _ := range common.DefaultModelRatio {
 		if strings.HasPrefix(model, "gpt-3.5") {
 			tokenEncoderMap[model] = gpt35TokenEncoder
 		} else if strings.HasPrefix(model, "gpt-4") {
@@ -116,6 +116,41 @@ func getImageToken(imageUrl *dto.MessageImageUrl) (int, error) {
 	return tiles*170 + 85, nil
 }
 
+func CountTokenChatRequest(request dto.GeneralOpenAIRequest, model string, checkSensitive bool) (int, error, bool) {
+	tkm := 0
+	msgTokens, err, b := CountTokenMessages(request.Messages, model, checkSensitive)
+	if err != nil {
+		return 0, err, b
+	}
+	tkm += msgTokens
+	if request.Tools != nil {
+		toolsData, _ := json.Marshal(request.Tools)
+		var openaiTools []dto.OpenAITools
+		err := json.Unmarshal(toolsData, &openaiTools)
+		if err != nil {
+			return 0, errors.New(fmt.Sprintf("count_tools_token_fail: %s", err.Error())), false
+		}
+		countStr := ""
+		for _, tool := range openaiTools {
+			countStr = tool.Function.Name
+			if tool.Function.Description != "" {
+				countStr += tool.Function.Description
+			}
+			if tool.Function.Parameters != nil {
+				countStr += fmt.Sprintf("%v", tool.Function.Parameters)
+			}
+		}
+		toolTokens, err, _ := CountTokenInput(countStr, model, false)
+		if err != nil {
+			return 0, err, false
+		}
+		tkm += 8
+		tkm += toolTokens
+	}
+
+	return tkm, nil, false
+}
+
 func CountTokenMessages(messages []dto.Message, model string, checkSensitive bool) (int, error, bool) {
 	//recover when panic
 	tokenEncoder := getTokenEncoder(model)
@@ -138,48 +173,31 @@ func CountTokenMessages(messages []dto.Message, model string, checkSensitive boo
 		tokenNum += tokensPerMessage
 		tokenNum += getTokenNum(tokenEncoder, message.Role)
 		if len(message.Content) > 0 {
-			var arrayContent []dto.MediaMessage
-			if err := json.Unmarshal(message.Content, &arrayContent); err != nil {
-				var stringContent string
-				if err := json.Unmarshal(message.Content, &stringContent); err != nil {
-					return 0, err, false
-				} else {
-					if checkSensitive {
-						contains, words := SensitiveWordContains(stringContent)
-						if contains {
-							err := fmt.Errorf("message contains sensitive words: [%s]", strings.Join(words, ", "))
-							return 0, err, true
-						}
-					}
-					tokenNum += getTokenNum(tokenEncoder, stringContent)
-					if message.Name != nil {
-						tokenNum += tokensPerName
-						tokenNum += getTokenNum(tokenEncoder, *message.Name)
+			if message.IsStringContent() {
+				stringContent := message.StringContent()
+				if checkSensitive {
+					contains, words := SensitiveWordContains(stringContent)
+					if contains {
+						err := fmt.Errorf("message contains sensitive words: [%s]", strings.Join(words, ", "))
+						return 0, err, true
 					}
 				}
+				tokenNum += getTokenNum(tokenEncoder, stringContent)
+				if message.Name != nil {
+					tokenNum += tokensPerName
+					tokenNum += getTokenNum(tokenEncoder, *message.Name)
+				}
 			} else {
+				var err error
+				arrayContent := message.ParseContent()
 				for _, m := range arrayContent {
 					if m.Type == "image_url" {
 						var imageTokenNum int
 						if model == "glm-4v" {
 							imageTokenNum = 1047
 						} else {
-							if str, ok := m.ImageUrl.(string); ok {
-								imageTokenNum, err = getImageToken(&dto.MessageImageUrl{Url: str, Detail: "auto"})
-							} else {
-								imageUrlMap := m.ImageUrl.(map[string]interface{})
-								detail, ok := imageUrlMap["detail"]
-								if ok {
-									imageUrlMap["detail"] = detail.(string)
-								} else {
-									imageUrlMap["detail"] = "auto"
-								}
-								imageUrl := dto.MessageImageUrl{
-									Url:    imageUrlMap["url"].(string),
-									Detail: imageUrlMap["detail"].(string),
-								}
-								imageTokenNum, err = getImageToken(&imageUrl)
-							}
+							imageUrl := m.ImageUrl.(dto.MessageImageUrl)
+							imageTokenNum, err = getImageToken(&imageUrl)
 							if err != nil {
 								return 0, err, false
 							}
@@ -208,7 +226,24 @@ func CountTokenInput(input any, model string, check bool) (int, error, bool) {
 		}
 		return CountTokenText(text, model, check)
 	}
-	return 0, errors.New("unsupported input type"), false
+	return CountTokenInput(fmt.Sprintf("%v", input), model, check)
+}
+
+func CountTokenStreamChoices(messages []dto.ChatCompletionsStreamResponseChoice, model string) int {
+	tokens := 0
+	for _, message := range messages {
+		tkm, _, _ := CountTokenInput(message.Delta.GetContentString(), model, false)
+		tokens += tkm
+		if message.Delta.ToolCalls != nil {
+			for _, tool := range message.Delta.ToolCalls {
+				tkm, _, _ := CountTokenInput(tool.Function.Name, model, false)
+				tokens += tkm
+				tkm, _, _ = CountTokenInput(tool.Function.Arguments, model, false)
+				tokens += tkm
+			}
+		}
+	}
+	return tokens
 }
 
 func CountAudioToken(text string, model string, check bool) (int, error, bool) {

web/.prettierrc.mjs

@@ -0,0 +1 @@
+module.exports = require("@so1ve/prettier-config");

web/README.md

@@ -18,4 +18,4 @@ Before you start editing, make sure your `Actions on Save` options have `Optimiz
 ## Reference
 
 1. https://github.com/OIerDb-ng/OIerDb
-2. https://github.com/cornflourblue/react-hooks-redux-registration-login-example
+2. https://github.com/cornflourblue/react-hooks-redux-registration-login-example

web/index.html

@@ -0,0 +1,19 @@
+<!doctype html>
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <link rel="icon" href="/logo.png" />
+    <meta name="viewport" content="width=device-width, initial-scale=1" />
+    <meta name="theme-color" content="#ffffff" />
+    <meta
+      name="description"
+      content="OpenAI 接口聚合管理，支持多种渠道包括 Azure，可用于二次分发管理 key，仅单可执行文件，已打包好 Docker 镜像，一键部署，开箱即用"
+    />
+    <title>New API</title>
+  </head>
+  <body>
+    <noscript>You need to enable JavaScript to run this app.</noscript>
+    <div id="root"></div>
+    <script type="module" src="/src/index.js"></script>
+  </body>
+</html>