Merge pull request #2815 from wans10/main

fix: 修复模型管理"参与官方同步"与"状态"开关无法保存的问题

common/topup-ratio.go

@@ -2,29 +2,37 @@ package common
 
 import (
 	"encoding/json"
+	"sync"
 )
 
-var TopupGroupRatio = map[string]float64{
+var topupGroupRatio = map[string]float64{
 	"default": 1,
 	"vip":     1,
 	"svip":    1,
 }
+var topupGroupRatioMutex sync.RWMutex
 
 func TopupGroupRatio2JSONString() string {
-	jsonBytes, err := json.Marshal(TopupGroupRatio)
+	topupGroupRatioMutex.RLock()
+	defer topupGroupRatioMutex.RUnlock()
+	jsonBytes, err := json.Marshal(topupGroupRatio)
 	if err != nil {
-		SysError("error marshalling model ratio: " + err.Error())
+		SysError("error marshalling topup group ratio: " + err.Error())
 	}
 	return string(jsonBytes)
 }
 
 func UpdateTopupGroupRatioByJSONString(jsonStr string) error {
-	TopupGroupRatio = make(map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &TopupGroupRatio)
+	topupGroupRatioMutex.Lock()
+	defer topupGroupRatioMutex.Unlock()
+	topupGroupRatio = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &topupGroupRatio)
 }
 
 func GetTopupGroupRatio(name string) float64 {
-	ratio, ok := TopupGroupRatio[name]
+	topupGroupRatioMutex.RLock()
+	defer topupGroupRatioMutex.RUnlock()
+	ratio, ok := topupGroupRatio[name]
 	if !ok {
 		SysError("topup group ratio not found: " + name)
 		return 1

controller/channel-test.go

@@ -31,6 +31,7 @@ import (
 
 	"github.com/bytedance/gopkg/util/gopool"
 	"github.com/samber/lo"
+	"github.com/tidwall/gjson"
 
 	"github.com/gin-gonic/gin"
 )
@@ -41,7 +42,21 @@ type testResult struct {
 	newAPIError *types.NewAPIError
 }
 
-func testChannel(channel *model.Channel, testModel string, endpointType string) testResult {
+func normalizeChannelTestEndpoint(channel *model.Channel, modelName, endpointType string) string {
+	normalized := strings.TrimSpace(endpointType)
+	if normalized != "" {
+		return normalized
+	}
+	if strings.HasSuffix(modelName, ratio_setting.CompactModelSuffix) {
+		return string(constant.EndpointTypeOpenAIResponseCompact)
+	}
+	if channel != nil && channel.Type == constant.ChannelTypeCodex {
+		return string(constant.EndpointTypeOpenAIResponse)
+	}
+	return normalized
+}
+
+func testChannel(channel *model.Channel, testModel string, endpointType string, isStream bool) testResult {
 	tik := time.Now()
 	var unsupportedTestChannelTypes = []int{
 		constant.ChannelTypeMidjourney,
@@ -76,6 +91,8 @@ func testChannel(channel *model.Channel, testModel string, endpointType string)
 		}
 	}
 
+	endpointType = normalizeChannelTestEndpoint(channel, testModel, endpointType)
+
 	requestPath := "/v1/chat/completions"
 
 	// 如果指定了端点类型，使用指定的端点类型
@@ -200,7 +217,7 @@ func testChannel(channel *model.Channel, testModel string, endpointType string)
 		}
 	}
 
-	request := buildTestRequest(testModel, endpointType, channel)
+	request := buildTestRequest(testModel, endpointType, channel, isStream)
 
 	info, err := relaycommon.GenRelayInfo(c, relayFormat, request, nil)
 
@@ -418,16 +435,16 @@ func testChannel(channel *model.Channel, testModel string, endpointType string)
 			newAPIError: respErr,
 		}
 	}
-	if usageA == nil {
+	usage, usageErr := coerceTestUsage(usageA, isStream, info.GetEstimatePromptTokens())
+	if usageErr != nil {
 		return testResult{
 			context:     c,
-			localErr:    errors.New("usage is nil"),
-			newAPIError: types.NewOpenAIError(errors.New("usage is nil"), types.ErrorCodeBadResponseBody, http.StatusInternalServerError),
+			localErr:    usageErr,
+			newAPIError: types.NewOpenAIError(usageErr, types.ErrorCodeBadResponseBody, http.StatusInternalServerError),
 		}
 	}
-	usage := usageA.(*dto.Usage)
 	result := w.Result()
-	respBody, err := io.ReadAll(result.Body)
+	respBody, err := readTestResponseBody(result.Body, isStream)
 	if err != nil {
 		return testResult{
 			context:     c,
@@ -435,6 +452,13 @@ func testChannel(channel *model.Channel, testModel string, endpointType string)
 			newAPIError: types.NewOpenAIError(err, types.ErrorCodeReadResponseBodyFailed, http.StatusInternalServerError),
 		}
 	}
+	if bodyErr := detectErrorFromTestResponseBody(respBody); bodyErr != nil {
+		return testResult{
+			context:     c,
+			localErr:    bodyErr,
+			newAPIError: types.NewOpenAIError(bodyErr, types.ErrorCodeBadResponseBody, http.StatusInternalServerError),
+		}
+	}
 	info.SetEstimatePromptTokens(usage.PromptTokens)
 
 	quota := 0
@@ -473,7 +497,101 @@ func testChannel(channel *model.Channel, testModel string, endpointType string)
 	}
 }
 
-func buildTestRequest(model string, endpointType string, channel *model.Channel) dto.Request {
+func coerceTestUsage(usageAny any, isStream bool, estimatePromptTokens int) (*dto.Usage, error) {
+	switch u := usageAny.(type) {
+	case *dto.Usage:
+		return u, nil
+	case dto.Usage:
+		return &u, nil
+	case nil:
+		if !isStream {
+			return nil, errors.New("usage is nil")
+		}
+		usage := &dto.Usage{
+			PromptTokens: estimatePromptTokens,
+		}
+		usage.TotalTokens = usage.PromptTokens
+		return usage, nil
+	default:
+		if !isStream {
+			return nil, fmt.Errorf("invalid usage type: %T", usageAny)
+		}
+		usage := &dto.Usage{
+			PromptTokens: estimatePromptTokens,
+		}
+		usage.TotalTokens = usage.PromptTokens
+		return usage, nil
+	}
+}
+
+func readTestResponseBody(body io.ReadCloser, isStream bool) ([]byte, error) {
+	defer func() { _ = body.Close() }()
+	const maxStreamLogBytes = 8 << 10
+	if isStream {
+		return io.ReadAll(io.LimitReader(body, maxStreamLogBytes))
+	}
+	return io.ReadAll(body)
+}
+
+func detectErrorFromTestResponseBody(respBody []byte) error {
+	b := bytes.TrimSpace(respBody)
+	if len(b) == 0 {
+		return nil
+	}
+	if message := detectErrorMessageFromJSONBytes(b); message != "" {
+		return fmt.Errorf("upstream error: %s", message)
+	}
+
+	for _, line := range bytes.Split(b, []byte{'\n'}) {
+		line = bytes.TrimSpace(line)
+		if len(line) == 0 {
+			continue
+		}
+		if !bytes.HasPrefix(line, []byte("data:")) {
+			continue
+		}
+		payload := bytes.TrimSpace(bytes.TrimPrefix(line, []byte("data:")))
+		if len(payload) == 0 || bytes.Equal(payload, []byte("[DONE]")) {
+			continue
+		}
+		if message := detectErrorMessageFromJSONBytes(payload); message != "" {
+			return fmt.Errorf("upstream error: %s", message)
+		}
+	}
+
+	return nil
+}
+
+func detectErrorMessageFromJSONBytes(jsonBytes []byte) string {
+	if len(jsonBytes) == 0 {
+		return ""
+	}
+	if jsonBytes[0] != '{' && jsonBytes[0] != '[' {
+		return ""
+	}
+	errVal := gjson.GetBytes(jsonBytes, "error")
+	if !errVal.Exists() || errVal.Type == gjson.Null {
+		return ""
+	}
+
+	message := gjson.GetBytes(jsonBytes, "error.message").String()
+	if message == "" {
+		message = gjson.GetBytes(jsonBytes, "error.error.message").String()
+	}
+	if message == "" && errVal.Type == gjson.String {
+		message = errVal.String()
+	}
+	if message == "" {
+		message = errVal.Raw
+	}
+	message = strings.TrimSpace(message)
+	if message == "" {
+		return "upstream returned error payload"
+	}
+	return message
+}
+
+func buildTestRequest(model string, endpointType string, channel *model.Channel, isStream bool) dto.Request {
 	testResponsesInput := json.RawMessage(`[{"role":"user","content":"hi"}]`)
 
 	// 根据端点类型构建不同的测试请求
@@ -504,8 +622,9 @@ func buildTestRequest(model string, endpointType string, channel *model.Channel)
 		case constant.EndpointTypeOpenAIResponse:
 			// 返回 OpenAIResponsesRequest
 			return &dto.OpenAIResponsesRequest{
-				Model: model,
-				Input: json.RawMessage(`[{"role":"user","content":"hi"}]`),
+				Model:  model,
+				Input:  json.RawMessage(`[{"role":"user","content":"hi"}]`),
+				Stream: isStream,
 			}
 		case constant.EndpointTypeOpenAIResponseCompact:
 			// 返回 OpenAIResponsesCompactionRequest
@@ -519,9 +638,9 @@ func buildTestRequest(model string, endpointType string, channel *model.Channel)
 			if constant.EndpointType(endpointType) == constant.EndpointTypeGemini {
 				maxTokens = 3000
 			}
-			return &dto.GeneralOpenAIRequest{
+			req := &dto.GeneralOpenAIRequest{
 				Model:  model,
-				Stream: false,
+				Stream: isStream,
 				Messages: []dto.Message{
 					{
 						Role:    "user",
@@ -530,6 +649,10 @@ func buildTestRequest(model string, endpointType string, channel *model.Channel)
 				},
 				MaxTokens: maxTokens,
 			}
+			if isStream {
+				req.StreamOptions = &dto.StreamOptions{IncludeUsage: true}
+			}
+			return req
 		}
 	}
 
@@ -565,15 +688,16 @@ func buildTestRequest(model string, endpointType string, channel *model.Channel)
 	// Responses-only models (e.g. codex series)
 	if strings.Contains(strings.ToLower(model), "codex") {
 		return &dto.OpenAIResponsesRequest{
-			Model: model,
-			Input: json.RawMessage(`[{"role":"user","content":"hi"}]`),
+			Model:  model,
+			Input:  json.RawMessage(`[{"role":"user","content":"hi"}]`),
+			Stream: isStream,
 		}
 	}
 
 	// Chat/Completion 请求 - 返回 GeneralOpenAIRequest
 	testRequest := &dto.GeneralOpenAIRequest{
 		Model:  model,
-		Stream: false,
+		Stream: isStream,
 		Messages: []dto.Message{
 			{
 				Role:    "user",
@@ -581,6 +705,9 @@ func buildTestRequest(model string, endpointType string, channel *model.Channel)
 			},
 		},
 	}
+	if isStream {
+		testRequest.StreamOptions = &dto.StreamOptions{IncludeUsage: true}
+	}
 
 	if strings.HasPrefix(model, "o") {
 		testRequest.MaxCompletionTokens = 16
@@ -618,8 +745,9 @@ func TestChannel(c *gin.Context) {
 	//}()
 	testModel := c.Query("model")
 	endpointType := c.Query("endpoint_type")
+	isStream, _ := strconv.ParseBool(c.Query("stream"))
 	tik := time.Now()
-	result := testChannel(channel, testModel, endpointType)
+	result := testChannel(channel, testModel, endpointType, isStream)
 	if result.localErr != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -678,7 +806,7 @@ func testAllChannels(notify bool) error {
 		for _, channel := range channels {
 			isChannelEnabled := channel.Status == common.ChannelStatusEnabled
 			tik := time.Now()
-			result := testChannel(channel, "", "")
+			result := testChannel(channel, "", "", false)
 			tok := time.Now()
 			milliseconds := tok.Sub(tik).Milliseconds()
 

controller/custom_oauth.go

@@ -166,21 +166,21 @@ func CreateCustomOAuthProvider(c *gin.Context) {
 
 // UpdateCustomOAuthProviderRequest is the request structure for updating a custom OAuth provider
 type UpdateCustomOAuthProviderRequest struct {
-	Name                  string `json:"name"`
-	Slug                  string `json:"slug"`
-	Enabled               bool   `json:"enabled"`
-	ClientId              string `json:"client_id"`
-	ClientSecret          string `json:"client_secret"` // Optional: if empty, keep existing
-	AuthorizationEndpoint string `json:"authorization_endpoint"`
-	TokenEndpoint         string `json:"token_endpoint"`
-	UserInfoEndpoint      string `json:"user_info_endpoint"`
-	Scopes                string `json:"scopes"`
-	UserIdField           string `json:"user_id_field"`
-	UsernameField         string `json:"username_field"`
-	DisplayNameField      string `json:"display_name_field"`
-	EmailField            string `json:"email_field"`
-	WellKnown             string `json:"well_known"`
-	AuthStyle             int    `json:"auth_style"`
+	Name                  string  `json:"name"`
+	Slug                  string  `json:"slug"`
+	Enabled               *bool   `json:"enabled"`               // Optional: if nil, keep existing
+	ClientId              string  `json:"client_id"`
+	ClientSecret          string  `json:"client_secret"`         // Optional: if empty, keep existing
+	AuthorizationEndpoint string  `json:"authorization_endpoint"`
+	TokenEndpoint         string  `json:"token_endpoint"`
+	UserInfoEndpoint      string  `json:"user_info_endpoint"`
+	Scopes                string  `json:"scopes"`
+	UserIdField           string  `json:"user_id_field"`
+	UsernameField         string  `json:"username_field"`
+	DisplayNameField      string  `json:"display_name_field"`
+	EmailField            string  `json:"email_field"`
+	WellKnown             *string `json:"well_known"`            // Optional: if nil, keep existing
+	AuthStyle             *int    `json:"auth_style"`            // Optional: if nil, keep existing
 }
 
 // UpdateCustomOAuthProvider updates an existing custom OAuth provider
@@ -227,7 +227,9 @@ func UpdateCustomOAuthProvider(c *gin.Context) {
 	if req.Slug != "" {
 		provider.Slug = req.Slug
 	}
-	provider.Enabled = req.Enabled
+	if req.Enabled != nil {
+		provider.Enabled = *req.Enabled
+	}
 	if req.ClientId != "" {
 		provider.ClientId = req.ClientId
 	}
@@ -258,8 +260,12 @@ func UpdateCustomOAuthProvider(c *gin.Context) {
 	if req.EmailField != "" {
 		provider.EmailField = req.EmailField
 	}
-	provider.WellKnown = req.WellKnown
-	provider.AuthStyle = req.AuthStyle
+	if req.WellKnown != nil {
+		provider.WellKnown = *req.WellKnown
+	}
+	if req.AuthStyle != nil {
+		provider.AuthStyle = *req.AuthStyle
+	}
 
 	if err := model.UpdateCustomOAuthProvider(provider); err != nil {
 		common.ApiError(c, err)
@@ -296,7 +302,12 @@ func DeleteCustomOAuthProvider(c *gin.Context) {
 	}
 
 	// Check if there are any user bindings
-	count, _ := model.GetBindingCountByProviderId(id)
+	count, err := model.GetBindingCountByProviderId(id)
+	if err != nil {
+		common.SysError("Failed to get binding count for provider " + strconv.Itoa(id) + ": " + err.Error())
+		common.ApiErrorMsg(c, "检查用户绑定时发生错误，请稍后重试")
+		return
+	}
 	if count > 0 {
 		common.ApiErrorMsg(c, "该 OAuth 提供商还有用户绑定，无法删除。请先解除所有用户绑定。")
 		return

controller/oauth.go

@@ -11,6 +11,7 @@ import (
 	"github.com/QuantumNous/new-api/oauth"
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
+	"gorm.io/gorm"
 )
 
 // providerParams returns map with Provider key for i18n templates
@@ -256,27 +257,62 @@ func findOrCreateOAuthUser(c *gin.Context, provider oauth.Provider, oauthUser *o
 		inviterId, _ = model.GetUserIdByAffCode(affCode.(string))
 	}
 
-	if err := user.Insert(inviterId); err != nil {
-		return nil, err
-	}
-
-	// For custom providers, create the binding after user is created
+	// Use transaction to ensure user creation and OAuth binding are atomic
 	if genericProvider, ok := provider.(*oauth.GenericOAuthProvider); ok {
-		binding := &model.UserOAuthBinding{
-			UserId:         user.Id,
-			ProviderId:     genericProvider.GetProviderId(),
-			ProviderUserId: oauthUser.ProviderUserID,
-		}
-		if err := model.CreateUserOAuthBinding(binding); err != nil {
-			common.SysError(fmt.Sprintf("[OAuth] Failed to create binding for user %d: %s", user.Id, err.Error()))
-			// Don't fail the registration, just log the error
+		// Custom provider: create user and binding in a transaction
+		err := model.DB.Transaction(func(tx *gorm.DB) error {
+			// Create user
+			if err := user.InsertWithTx(tx, inviterId); err != nil {
+				return err
+			}
+
+			// Create OAuth binding
+			binding := &model.UserOAuthBinding{
+				UserId:         user.Id,
+				ProviderId:     genericProvider.GetProviderId(),
+				ProviderUserId: oauthUser.ProviderUserID,
+			}
+			if err := model.CreateUserOAuthBindingWithTx(tx, binding); err != nil {
+				return err
+			}
+
+			return nil
+		})
+		if err != nil {
+			return nil, err
 		}
+
+		// Perform post-transaction tasks (logs, sidebar config, inviter rewards)
+		user.FinalizeOAuthUserCreation(inviterId)
 	} else {
-		// Built-in provider: set the provider user ID on the user model
-		provider.SetProviderUserID(user, oauthUser.ProviderUserID)
-		if err := user.Update(false); err != nil {
-			common.SysError(fmt.Sprintf("[OAuth] Failed to update provider ID for user %d: %s", user.Id, err.Error()))
+		// Built-in provider: create user and update provider ID in a transaction
+		err := model.DB.Transaction(func(tx *gorm.DB) error {
+			// Create user
+			if err := user.InsertWithTx(tx, inviterId); err != nil {
+				return err
+			}
+
+			// Set the provider user ID on the user model and update
+			provider.SetProviderUserID(user, oauthUser.ProviderUserID)
+			if err := tx.Model(user).Updates(map[string]interface{}{
+				"github_id":    user.GitHubId,
+				"discord_id":   user.DiscordId,
+				"oidc_id":      user.OidcId,
+				"linux_do_id":  user.LinuxDOId,
+				"wechat_id":    user.WeChatId,
+				"telegram_id":  user.TelegramId,
+			}).Error; err != nil {
+				return err
+			}
+
+			return nil
+		})
+		if err != nil {
+			return nil, err
 		}
+
+		// Perform post-transaction tasks
+		user.FinalizeOAuthUserCreation(inviterId)
 	}
 
 	return user, nil

controller/option.go

@@ -169,6 +169,15 @@ func UpdateOption(c *gin.Context) {
 			})
 			return
 		}
+	case "CreateCacheRatio":
+		err = ratio_setting.UpdateCreateCacheRatioByJSONString(option.Value.(string))
+		if err != nil {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "缓存创建倍率设置失败: " + err.Error(),
+			})
+			return
+		}
 	case "ModelRequestRateLimitGroup":
 		err = setting.CheckModelRequestRateLimitGroup(option.Value.(string))
 		if err != nil {

dto/channel_settings.go

@@ -27,6 +27,7 @@ type ChannelOtherSettings struct {
 	AzureResponsesVersion string        `json:"azure_responses_version,omitempty"`
 	VertexKeyType         VertexKeyType `json:"vertex_key_type,omitempty"` // "json" or "api_key"
 	OpenRouterEnterprise  *bool         `json:"openrouter_enterprise,omitempty"`
+	ClaudeBetaQuery       bool          `json:"claude_beta_query,omitempty"`      // Claude 渠道是否强制追加 ?beta=true
 	AllowServiceTier      bool          `json:"allow_service_tier,omitempty"`      // 是否允许 service_tier 透传（默认过滤以避免额外计费）
 	DisableStore          bool          `json:"disable_store,omitempty"`           // 是否禁用 store 透传（默认允许透传，禁用后可能导致 Codex 无法使用）
 	AllowSafetyIdentifier bool          `json:"allow_safety_identifier,omitempty"` // 是否允许 safety_identifier 透传（默认过滤以保护用户隐私）

model/custom_oauth_provider.go

@@ -97,13 +97,18 @@ func DeleteCustomOAuthProvider(id int) error {
 }
 
 // IsSlugTaken checks if a slug is already taken by another provider
+// Returns true on DB errors (fail-closed) to prevent slug conflicts
 func IsSlugTaken(slug string, excludeId int) bool {
 	var count int64
 	query := DB.Model(&CustomOAuthProvider{}).Where("slug = ?", slug)
 	if excludeId > 0 {
 		query = query.Where("id != ?", excludeId)
 	}
-	query.Count(&count)
+	res := query.Count(&count)
+	if res.Error != nil {
+		// Fail-closed: treat DB errors as slug being taken to prevent conflicts
+		return true
+	}
 	return count > 0
 }
 

model/model_meta.go

@@ -47,7 +47,21 @@ func (mi *Model) Insert() error {
 	now := common.GetTimestamp()
 	mi.CreatedTime = now
 	mi.UpdatedTime = now
-	return DB.Create(mi).Error
+
+	// 保存原始值（因为 Create 后可能被 GORM 的 default 标签覆盖为 1）
+	originalStatus := mi.Status
+	originalSyncOfficial := mi.SyncOfficial
+
+	// 先创建记录（GORM 会对零值字段应用默认值）
+	if err := DB.Create(mi).Error; err != nil {
+		return err
+	}
+
+	// 使用保存的原始值进行更新，确保零值能正确保存
+	return DB.Model(&Model{}).Where("id = ?", mi.Id).Updates(map[string]interface{}{
+		"status":        originalStatus,
+		"sync_official": originalSyncOfficial,
+	}).Error
 }
 
 func IsModelNameDuplicated(id int, name string) (bool, error) {
@@ -61,11 +75,9 @@ func IsModelNameDuplicated(id int, name string) (bool, error) {
 
 func (mi *Model) Update() error {
 	mi.UpdatedTime = common.GetTimestamp()
-	return DB.Session(&gorm.Session{AllowGlobalUpdate: false, FullSaveAssociations: false}).
-		Model(&Model{}).
-		Where("id = ?", mi.Id).
-		Omit("created_time").
-		Select("*").
+	// 使用 Select 强制更新所有字段，包括零值
+	return DB.Model(&Model{}).Where("id = ?", mi.Id).
+		Select("model_name", "description", "icon", "tags", "vendor_id", "endpoints", "status", "sync_official", "name_rule", "updated_time").
 		Updates(mi).Error
 }
 

model/option.go

@@ -115,6 +115,7 @@ func InitOptionMap() {
 	common.OptionMap["ModelRatio"] = ratio_setting.ModelRatio2JSONString()
 	common.OptionMap["ModelPrice"] = ratio_setting.ModelPrice2JSONString()
 	common.OptionMap["CacheRatio"] = ratio_setting.CacheRatio2JSONString()
+	common.OptionMap["CreateCacheRatio"] = ratio_setting.CreateCacheRatio2JSONString()
 	common.OptionMap["GroupRatio"] = ratio_setting.GroupRatio2JSONString()
 	common.OptionMap["GroupGroupRatio"] = ratio_setting.GroupGroupRatio2JSONString()
 	common.OptionMap["UserUsableGroups"] = setting.UserUsableGroups2JSONString()
@@ -427,6 +428,8 @@ func updateOptionMap(key string, value string) (err error) {
 		err = ratio_setting.UpdateModelPriceByJSONString(value)
 	case "CacheRatio":
 		err = ratio_setting.UpdateCacheRatioByJSONString(value)
+	case "CreateCacheRatio":
+		err = ratio_setting.UpdateCreateCacheRatioByJSONString(value)
 	case "ImageRatio":
 		err = ratio_setting.UpdateImageRatioByJSONString(value)
 	case "AudioRatio":

model/pricing.go

@@ -196,20 +196,25 @@ func updatePricing() {
 		modelSupportEndpointsStr[ability.Model] = endpoints
 	}
 
-	// 再补充模型自定义端点
+	// 再补充模型自定义端点：若配置有效则替换默认端点，不做合并
 	for modelName, meta := range metaMap {
 		if strings.TrimSpace(meta.Endpoints) == "" {
 			continue
 		}
 		var raw map[string]interface{}
 		if err := json.Unmarshal([]byte(meta.Endpoints), &raw); err == nil {
-			endpoints := modelSupportEndpointsStr[modelName]
-			for k := range raw {
-				if !common.StringsContains(endpoints, k) {
-					endpoints = append(endpoints, k)
+			endpoints := make([]string, 0, len(raw))
+			for k, v := range raw {
+				switch v.(type) {
+				case string, map[string]interface{}:
+					if !common.StringsContains(endpoints, k) {
+						endpoints = append(endpoints, k)
+					}
 				}
 			}
-			modelSupportEndpointsStr[modelName] = endpoints
+			if len(endpoints) > 0 {
+				modelSupportEndpointsStr[modelName] = endpoints
+			}
 		}
 	}
 

model/user.go

@@ -429,6 +429,65 @@ func (user *User) Insert(inviterId int) error {
 	return nil
 }
 
+// InsertWithTx inserts a new user within an existing transaction.
+// This is used for OAuth registration where user creation and binding need to be atomic.
+// Post-creation tasks (sidebar config, logs, inviter rewards) are handled after the transaction commits.
+func (user *User) InsertWithTx(tx *gorm.DB, inviterId int) error {
+	var err error
+	if user.Password != "" {
+		user.Password, err = common.Password2Hash(user.Password)
+		if err != nil {
+			return err
+		}
+	}
+	user.Quota = common.QuotaForNewUser
+	user.AffCode = common.GetRandomString(4)
+
+	// 初始化用户设置
+	if user.Setting == "" {
+		defaultSetting := dto.UserSetting{}
+		user.SetSetting(defaultSetting)
+	}
+
+	result := tx.Create(user)
+	if result.Error != nil {
+		return result.Error
+	}
+
+	return nil
+}
+
+// FinalizeOAuthUserCreation performs post-transaction tasks for OAuth user creation.
+// This should be called after the transaction commits successfully.
+func (user *User) FinalizeOAuthUserCreation(inviterId int) {
+	// 用户创建成功后，根据角色初始化边栏配置
+	var createdUser User
+	if err := DB.Where("id = ?", user.Id).First(&createdUser).Error; err == nil {
+		defaultSidebarConfig := generateDefaultSidebarConfigForRole(createdUser.Role)
+		if defaultSidebarConfig != "" {
+			currentSetting := createdUser.GetSetting()
+			currentSetting.SidebarModules = defaultSidebarConfig
+			createdUser.SetSetting(currentSetting)
+			createdUser.Update(false)
+			common.SysLog(fmt.Sprintf("为新用户 %s (角色: %d) 初始化边栏配置", createdUser.Username, createdUser.Role))
+		}
+	}
+
+	if common.QuotaForNewUser > 0 {
+		RecordLog(user.Id, LogTypeSystem, fmt.Sprintf("新用户注册赠送 %s", logger.LogQuota(common.QuotaForNewUser)))
+	}
+	if inviterId != 0 {
+		if common.QuotaForInvitee > 0 {
+			_ = IncreaseUserQuota(user.Id, common.QuotaForInvitee, true)
+			RecordLog(user.Id, LogTypeSystem, fmt.Sprintf("使用邀请码赠送 %s", logger.LogQuota(common.QuotaForInvitee)))
+		}
+		if common.QuotaForInviter > 0 {
+			RecordLog(inviterId, LogTypeSystem, fmt.Sprintf("邀请用户赠送 %s", logger.LogQuota(common.QuotaForInviter)))
+			_ = inviteUser(inviterId)
+		}
+	}
+}
+
 func (user *User) Update(updatePassword bool) error {
 	var err error
 	if updatePassword {

model/user_oauth_binding.go

@@ -3,18 +3,17 @@ package model
 import (
 	"errors"
 	"time"
+
+	"gorm.io/gorm"
 )
 
 // UserOAuthBinding stores the binding relationship between users and custom OAuth providers
 type UserOAuthBinding struct {
 	Id             int       `json:"id" gorm:"primaryKey"`
-	UserId         int       `json:"user_id" gorm:"index;not null"`                                               // User ID
-	ProviderId     int       `json:"provider_id" gorm:"index;not null"`                                           // Custom OAuth provider ID
-	ProviderUserId string    `json:"provider_user_id" gorm:"type:varchar(256);not null"`                          // User ID from OAuth provider
+	UserId         int       `json:"user_id" gorm:"not null;uniqueIndex:ux_user_provider"`                                        // User ID - one binding per user per provider
+	ProviderId     int       `json:"provider_id" gorm:"not null;uniqueIndex:ux_user_provider;uniqueIndex:ux_provider_userid"`     // Custom OAuth provider ID
+	ProviderUserId string    `json:"provider_user_id" gorm:"type:varchar(256);not null;uniqueIndex:ux_provider_userid"`           // User ID from OAuth provider - one OAuth account per provider
 	CreatedAt      time.Time `json:"created_at"`
-
-	// Composite unique index to prevent duplicate bindings
-	// One OAuth account can only be bound to one user
 }
 
 func (UserOAuthBinding) TableName() string {
@@ -82,6 +81,29 @@ func CreateUserOAuthBinding(binding *UserOAuthBinding) error {
 	return DB.Create(binding).Error
 }
 
+// CreateUserOAuthBindingWithTx creates a new OAuth binding within a transaction
+func CreateUserOAuthBindingWithTx(tx *gorm.DB, binding *UserOAuthBinding) error {
+	if binding.UserId == 0 {
+		return errors.New("user ID is required")
+	}
+	if binding.ProviderId == 0 {
+		return errors.New("provider ID is required")
+	}
+	if binding.ProviderUserId == "" {
+		return errors.New("provider user ID is required")
+	}
+
+	// Check if this provider user ID is already taken (use tx to check within the same transaction)
+	var count int64
+	tx.Model(&UserOAuthBinding{}).Where("provider_id = ? AND provider_user_id = ?", binding.ProviderId, binding.ProviderUserId).Count(&count)
+	if count > 0 {
+		return errors.New("this OAuth account is already bound to another user")
+	}
+
+	binding.CreatedAt = time.Now()
+	return tx.Create(binding).Error
+}
+
 // UpdateUserOAuthBinding updates an existing OAuth binding (e.g., rebind to different OAuth account)
 func UpdateUserOAuthBinding(userId, providerId int, newProviderUserId string) error {
 	// Check if the new provider user ID is already taken by another user

oauth/github.go

@@ -5,6 +5,7 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
+	"io"
 	"net/http"
 	"strconv"
 	"time"
@@ -122,6 +123,17 @@ func (p *GitHubProvider) GetUserInfo(ctx context.Context, token *OAuthToken) (*O
 
 	logger.LogDebug(ctx, "[OAuth-GitHub] GetUserInfo response status: %d", res.StatusCode)
 
+	// Check for non-200 status codes before attempting to decode
+	if res.StatusCode != http.StatusOK {
+		body, _ := io.ReadAll(res.Body)
+		bodyStr := string(body)
+		if len(bodyStr) > 500 {
+			bodyStr = bodyStr[:500] + "..."
+		}
+		logger.LogError(ctx, fmt.Sprintf("[OAuth-GitHub] GetUserInfo failed: status=%d, body=%s", res.StatusCode, bodyStr))
+		return nil, NewOAuthErrorWithRaw(i18n.MsgOAuthGetUserErr, map[string]any{"Provider": "GitHub"}, fmt.Sprintf("status %d", res.StatusCode))
+	}
+
 	var githubUser gitHubUser
 	err = json.NewDecoder(res.Body).Decode(&githubUser)
 	if err != nil {

relay/channel/ali/adaptor.go

@@ -223,11 +223,8 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 	switch info.RelayFormat {
 	case types.RelayFormatClaude:
 		if supportsAliAnthropicMessages(info.UpstreamModelName) {
-			if info.IsStream {
-				return claude.ClaudeStreamHandler(c, resp, info)
-			}
-
-			return claude.ClaudeHandler(c, resp, info)
+			adaptor := claude.Adaptor{}
+			return adaptor.DoResponse(c, resp, info)
 		}
 
 		adaptor := openai.Adaptor{}

relay/channel/claude/adaptor.go

@@ -95,6 +95,7 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 }
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
+	info.FinalRequestRelayFormat = types.RelayFormatClaude
 	if info.IsStream {
 		return ClaudeStreamHandler(c, resp, info)
 	} else {

relay/channel/claude/message_delta_usage_patch_test.go

@@ -0,0 +1,111 @@
+package claude
+
+import (
+	"testing"
+
+	"github.com/QuantumNous/new-api/dto"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	"github.com/QuantumNous/new-api/setting/model_setting"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"github.com/tidwall/gjson"
+)
+
+func TestPatchClaudeMessageDeltaUsageDataPreserveUnknownFields(t *testing.T) {
+	originalData := `{"type":"message_delta","delta":{"stop_reason":"end_turn","stop_sequence":null},"usage":{"output_tokens":53},"vendor_meta":{"trace_id":"trace_001"}}`
+	usage := &dto.ClaudeUsage{
+		InputTokens:              100,
+		CacheReadInputTokens:     30,
+		CacheCreationInputTokens: 50,
+	}
+
+	patchedData := patchClaudeMessageDeltaUsageData(originalData, usage)
+
+	require.Equal(t, "message_delta", gjson.Get(patchedData, "type").String())
+	require.Equal(t, "end_turn", gjson.Get(patchedData, "delta.stop_reason").String())
+	require.Equal(t, "trace_001", gjson.Get(patchedData, "vendor_meta.trace_id").String())
+	require.EqualValues(t, 53, gjson.Get(patchedData, "usage.output_tokens").Int())
+	require.EqualValues(t, 100, gjson.Get(patchedData, "usage.input_tokens").Int())
+	require.EqualValues(t, 30, gjson.Get(patchedData, "usage.cache_read_input_tokens").Int())
+	require.EqualValues(t, 50, gjson.Get(patchedData, "usage.cache_creation_input_tokens").Int())
+}
+
+func TestPatchClaudeMessageDeltaUsageDataZeroValueChecks(t *testing.T) {
+	originalData := `{"type":"message_delta","usage":{"output_tokens":53,"input_tokens":9,"cache_read_input_tokens":0}}`
+	usage := &dto.ClaudeUsage{
+		InputTokens:              100,
+		CacheReadInputTokens:     30,
+		CacheCreationInputTokens: 0,
+	}
+
+	patchedData := patchClaudeMessageDeltaUsageData(originalData, usage)
+
+	require.EqualValues(t, 9, gjson.Get(patchedData, "usage.input_tokens").Int())
+	require.EqualValues(t, 30, gjson.Get(patchedData, "usage.cache_read_input_tokens").Int())
+	assert.False(t, gjson.Get(patchedData, "usage.cache_creation_input_tokens").Exists())
+}
+
+func TestShouldSkipClaudeMessageDeltaUsagePatch(t *testing.T) {
+	originGlobalPassThrough := model_setting.GetGlobalSettings().PassThroughRequestEnabled
+	t.Cleanup(func() {
+		model_setting.GetGlobalSettings().PassThroughRequestEnabled = originGlobalPassThrough
+	})
+
+	model_setting.GetGlobalSettings().PassThroughRequestEnabled = true
+	assert.True(t, shouldSkipClaudeMessageDeltaUsagePatch(&relaycommon.RelayInfo{}))
+
+	model_setting.GetGlobalSettings().PassThroughRequestEnabled = false
+	assert.True(t, shouldSkipClaudeMessageDeltaUsagePatch(&relaycommon.RelayInfo{
+		ChannelMeta: &relaycommon.ChannelMeta{ChannelSetting: dto.ChannelSettings{PassThroughBodyEnabled: true}},
+	}))
+	assert.False(t, shouldSkipClaudeMessageDeltaUsagePatch(&relaycommon.RelayInfo{
+		ChannelMeta: &relaycommon.ChannelMeta{ChannelSetting: dto.ChannelSettings{PassThroughBodyEnabled: false}},
+	}))
+}
+
+func TestBuildMessageDeltaPatchUsage(t *testing.T) {
+	t.Run("merge missing fields from claudeInfo", func(t *testing.T) {
+		claudeResponse := &dto.ClaudeResponse{Usage: &dto.ClaudeUsage{OutputTokens: 53}}
+		claudeInfo := &ClaudeResponseInfo{
+			Usage: &dto.Usage{
+				PromptTokens: 100,
+				PromptTokensDetails: dto.InputTokenDetails{
+					CachedTokens:         30,
+					CachedCreationTokens: 50,
+				},
+				ClaudeCacheCreation5mTokens: 10,
+				ClaudeCacheCreation1hTokens: 20,
+			},
+		}
+
+		usage := buildMessageDeltaPatchUsage(claudeResponse, claudeInfo)
+		require.NotNil(t, usage)
+		require.EqualValues(t, 100, usage.InputTokens)
+		require.EqualValues(t, 30, usage.CacheReadInputTokens)
+		require.EqualValues(t, 50, usage.CacheCreationInputTokens)
+		require.EqualValues(t, 53, usage.OutputTokens)
+		require.NotNil(t, usage.CacheCreation)
+		require.EqualValues(t, 10, usage.CacheCreation.Ephemeral5mInputTokens)
+		require.EqualValues(t, 20, usage.CacheCreation.Ephemeral1hInputTokens)
+	})
+
+	t.Run("keep upstream non-zero values", func(t *testing.T) {
+		claudeResponse := &dto.ClaudeResponse{Usage: &dto.ClaudeUsage{
+			InputTokens:              9,
+			CacheReadInputTokens:     7,
+			CacheCreationInputTokens: 6,
+		}}
+		claudeInfo := &ClaudeResponseInfo{Usage: &dto.Usage{
+			PromptTokens: 100,
+			PromptTokensDetails: dto.InputTokenDetails{
+				CachedTokens:         30,
+				CachedCreationTokens: 50,
+			},
+		}}
+
+		usage := buildMessageDeltaPatchUsage(claudeResponse, claudeInfo)
+		require.EqualValues(t, 9, usage.InputTokens)
+		require.EqualValues(t, 7, usage.CacheReadInputTokens)
+		require.EqualValues(t, 6, usage.CacheCreationInputTokens)
+	})
+}

relay/channel/claude/relay-claude.go

@@ -21,6 +21,8 @@ import (
 	"github.com/QuantumNous/new-api/types"
 
 	"github.com/gin-gonic/gin"
+	"github.com/tidwall/gjson"
+	"github.com/tidwall/sjson"
 )
 
 const (
@@ -544,6 +546,78 @@ type ClaudeResponseInfo struct {
 	Done         bool
 }
 
+func buildMessageDeltaPatchUsage(claudeResponse *dto.ClaudeResponse, claudeInfo *ClaudeResponseInfo) *dto.ClaudeUsage {
+	usage := &dto.ClaudeUsage{}
+	if claudeResponse != nil && claudeResponse.Usage != nil {
+		*usage = *claudeResponse.Usage
+	}
+
+	if claudeInfo == nil || claudeInfo.Usage == nil {
+		return usage
+	}
+
+	if usage.InputTokens == 0 && claudeInfo.Usage.PromptTokens > 0 {
+		usage.InputTokens = claudeInfo.Usage.PromptTokens
+	}
+	if usage.CacheReadInputTokens == 0 && claudeInfo.Usage.PromptTokensDetails.CachedTokens > 0 {
+		usage.CacheReadInputTokens = claudeInfo.Usage.PromptTokensDetails.CachedTokens
+	}
+	if usage.CacheCreationInputTokens == 0 && claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens > 0 {
+		usage.CacheCreationInputTokens = claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens
+	}
+	if usage.CacheCreation == nil && (claudeInfo.Usage.ClaudeCacheCreation5mTokens > 0 || claudeInfo.Usage.ClaudeCacheCreation1hTokens > 0) {
+		usage.CacheCreation = &dto.ClaudeCacheCreationUsage{
+			Ephemeral5mInputTokens: claudeInfo.Usage.ClaudeCacheCreation5mTokens,
+			Ephemeral1hInputTokens: claudeInfo.Usage.ClaudeCacheCreation1hTokens,
+		}
+	}
+	return usage
+}
+
+func shouldSkipClaudeMessageDeltaUsagePatch(info *relaycommon.RelayInfo) bool {
+	if model_setting.GetGlobalSettings().PassThroughRequestEnabled {
+		return true
+	}
+	if info == nil {
+		return false
+	}
+	return info.ChannelSetting.PassThroughBodyEnabled
+}
+
+func patchClaudeMessageDeltaUsageData(data string, usage *dto.ClaudeUsage) string {
+	if data == "" || usage == nil {
+		return data
+	}
+
+	data = setMessageDeltaUsageInt(data, "usage.input_tokens", usage.InputTokens)
+	data = setMessageDeltaUsageInt(data, "usage.cache_read_input_tokens", usage.CacheReadInputTokens)
+	data = setMessageDeltaUsageInt(data, "usage.cache_creation_input_tokens", usage.CacheCreationInputTokens)
+
+	if usage.CacheCreation != nil {
+		data = setMessageDeltaUsageInt(data, "usage.cache_creation.ephemeral_5m_input_tokens", usage.CacheCreation.Ephemeral5mInputTokens)
+		data = setMessageDeltaUsageInt(data, "usage.cache_creation.ephemeral_1h_input_tokens", usage.CacheCreation.Ephemeral1hInputTokens)
+	}
+
+	return data
+}
+
+func setMessageDeltaUsageInt(data string, path string, localValue int) string {
+	if localValue <= 0 {
+		return data
+	}
+
+	upstreamValue := gjson.Get(data, path)
+	if upstreamValue.Exists() && upstreamValue.Int() > 0 {
+		return data
+	}
+
+	patchedData, err := sjson.Set(data, path, localValue)
+	if err != nil {
+		return data
+	}
+	return patchedData
+}
+
 func FormatClaudeResponseInfo(claudeResponse *dto.ClaudeResponse, oaiResponse *dto.ChatCompletionsStreamResponse, claudeInfo *ClaudeResponseInfo) bool {
 	if claudeInfo == nil {
 		return false
@@ -638,6 +712,12 @@ func HandleStreamResponseData(c *gin.Context, info *relaycommon.RelayInfo, claud
 			if claudeResponse.Message != nil {
 				info.UpstreamModelName = claudeResponse.Message.Model
 			}
+		} else if claudeResponse.Type == "message_delta" {
+			// 确保 message_delta 的 usage 包含完整的 input_tokens 和 cache 相关字段
+			// 解决 AWS Bedrock 等上游返回的 message_delta 缺少这些字段的问题
+			if !shouldSkipClaudeMessageDeltaUsagePatch(info) {
+				data = patchClaudeMessageDeltaUsageData(data, buildMessageDeltaPatchUsage(&claudeResponse, claudeInfo))
+			}
 		}
 		helper.ClaudeChunkData(c, claudeResponse, data)
 	} else if info.RelayFormat == types.RelayFormatOpenAI {

relay/channel/claude/relay_claude_test.go

@@ -0,0 +1,175 @@
+package claude
+
+import (
+	"strings"
+	"testing"
+
+	"github.com/QuantumNous/new-api/dto"
+)
+
+func TestFormatClaudeResponseInfo_MessageStart(t *testing.T) {
+	claudeInfo := &ClaudeResponseInfo{
+		Usage: &dto.Usage{},
+	}
+	claudeResponse := &dto.ClaudeResponse{
+		Type: "message_start",
+		Message: &dto.ClaudeMediaMessage{
+			Id:    "msg_123",
+			Model: "claude-3-5-sonnet",
+			Usage: &dto.ClaudeUsage{
+				InputTokens:              100,
+				OutputTokens:             1,
+				CacheCreationInputTokens: 50,
+				CacheReadInputTokens:     30,
+			},
+		},
+	}
+
+	ok := FormatClaudeResponseInfo(claudeResponse, nil, claudeInfo)
+	if !ok {
+		t.Fatal("expected true")
+	}
+	if claudeInfo.Usage.PromptTokens != 100 {
+		t.Errorf("PromptTokens = %d, want 100", claudeInfo.Usage.PromptTokens)
+	}
+	if claudeInfo.Usage.PromptTokensDetails.CachedTokens != 30 {
+		t.Errorf("CachedTokens = %d, want 30", claudeInfo.Usage.PromptTokensDetails.CachedTokens)
+	}
+	if claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens != 50 {
+		t.Errorf("CachedCreationTokens = %d, want 50", claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens)
+	}
+	if claudeInfo.ResponseId != "msg_123" {
+		t.Errorf("ResponseId = %s, want msg_123", claudeInfo.ResponseId)
+	}
+	if claudeInfo.Model != "claude-3-5-sonnet" {
+		t.Errorf("Model = %s, want claude-3-5-sonnet", claudeInfo.Model)
+	}
+}
+
+func TestFormatClaudeResponseInfo_MessageDelta_FullUsage(t *testing.T) {
+	// message_start 先积累 usage
+	claudeInfo := &ClaudeResponseInfo{
+		Usage: &dto.Usage{
+			PromptTokens: 100,
+			PromptTokensDetails: dto.InputTokenDetails{
+				CachedTokens:         30,
+				CachedCreationTokens: 50,
+			},
+			CompletionTokens: 1,
+		},
+	}
+
+	// message_delta 带完整 usage（原生 Anthropic 场景）
+	claudeResponse := &dto.ClaudeResponse{
+		Type: "message_delta",
+		Usage: &dto.ClaudeUsage{
+			InputTokens:              100,
+			OutputTokens:             200,
+			CacheCreationInputTokens: 50,
+			CacheReadInputTokens:     30,
+		},
+	}
+
+	ok := FormatClaudeResponseInfo(claudeResponse, nil, claudeInfo)
+	if !ok {
+		t.Fatal("expected true")
+	}
+	if claudeInfo.Usage.PromptTokens != 100 {
+		t.Errorf("PromptTokens = %d, want 100", claudeInfo.Usage.PromptTokens)
+	}
+	if claudeInfo.Usage.CompletionTokens != 200 {
+		t.Errorf("CompletionTokens = %d, want 200", claudeInfo.Usage.CompletionTokens)
+	}
+	if claudeInfo.Usage.TotalTokens != 300 {
+		t.Errorf("TotalTokens = %d, want 300", claudeInfo.Usage.TotalTokens)
+	}
+	if !claudeInfo.Done {
+		t.Error("expected Done = true")
+	}
+}
+
+func TestFormatClaudeResponseInfo_MessageDelta_OnlyOutputTokens(t *testing.T) {
+	// 模拟 Bedrock: message_start 已积累 usage
+	claudeInfo := &ClaudeResponseInfo{
+		Usage: &dto.Usage{
+			PromptTokens: 100,
+			PromptTokensDetails: dto.InputTokenDetails{
+				CachedTokens:         30,
+				CachedCreationTokens: 50,
+			},
+			CompletionTokens:            1,
+			ClaudeCacheCreation5mTokens: 10,
+			ClaudeCacheCreation1hTokens: 20,
+		},
+	}
+
+	// Bedrock 的 message_delta 只有 output_tokens，缺少 input_tokens 和 cache 字段
+	claudeResponse := &dto.ClaudeResponse{
+		Type: "message_delta",
+		Usage: &dto.ClaudeUsage{
+			OutputTokens: 200,
+			// InputTokens, CacheCreationInputTokens, CacheReadInputTokens 都是 0
+		},
+	}
+
+	ok := FormatClaudeResponseInfo(claudeResponse, nil, claudeInfo)
+	if !ok {
+		t.Fatal("expected true")
+	}
+	// PromptTokens 应保持 message_start 的值（因为 message_delta 的 InputTokens=0，不更新）
+	if claudeInfo.Usage.PromptTokens != 100 {
+		t.Errorf("PromptTokens = %d, want 100", claudeInfo.Usage.PromptTokens)
+	}
+	if claudeInfo.Usage.CompletionTokens != 200 {
+		t.Errorf("CompletionTokens = %d, want 200", claudeInfo.Usage.CompletionTokens)
+	}
+	if claudeInfo.Usage.TotalTokens != 300 {
+		t.Errorf("TotalTokens = %d, want 300", claudeInfo.Usage.TotalTokens)
+	}
+	// cache 字段应保持 message_start 的值
+	if claudeInfo.Usage.PromptTokensDetails.CachedTokens != 30 {
+		t.Errorf("CachedTokens = %d, want 30", claudeInfo.Usage.PromptTokensDetails.CachedTokens)
+	}
+	if claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens != 50 {
+		t.Errorf("CachedCreationTokens = %d, want 50", claudeInfo.Usage.PromptTokensDetails.CachedCreationTokens)
+	}
+	if claudeInfo.Usage.ClaudeCacheCreation5mTokens != 10 {
+		t.Errorf("ClaudeCacheCreation5mTokens = %d, want 10", claudeInfo.Usage.ClaudeCacheCreation5mTokens)
+	}
+	if claudeInfo.Usage.ClaudeCacheCreation1hTokens != 20 {
+		t.Errorf("ClaudeCacheCreation1hTokens = %d, want 20", claudeInfo.Usage.ClaudeCacheCreation1hTokens)
+	}
+	if !claudeInfo.Done {
+		t.Error("expected Done = true")
+	}
+}
+
+func TestFormatClaudeResponseInfo_NilClaudeInfo(t *testing.T) {
+	claudeResponse := &dto.ClaudeResponse{Type: "message_start"}
+	ok := FormatClaudeResponseInfo(claudeResponse, nil, nil)
+	if ok {
+		t.Error("expected false for nil claudeInfo")
+	}
+}
+
+func TestFormatClaudeResponseInfo_ContentBlockDelta(t *testing.T) {
+	text := "hello"
+	claudeInfo := &ClaudeResponseInfo{
+		Usage:        &dto.Usage{},
+		ResponseText: strings.Builder{},
+	}
+	claudeResponse := &dto.ClaudeResponse{
+		Type: "content_block_delta",
+		Delta: &dto.ClaudeMediaMessage{
+			Text: &text,
+		},
+	}
+
+	ok := FormatClaudeResponseInfo(claudeResponse, nil, claudeInfo)
+	if !ok {
+		t.Fatal("expected true")
+	}
+	if claudeInfo.ResponseText.String() != "hello" {
+		t.Errorf("ResponseText = %q, want %q", claudeInfo.ResponseText.String(), "hello")
+	}
+}

relay/channel/codex/adaptor.go

@@ -26,7 +26,7 @@ func (a *Adaptor) ConvertGeminiRequest(c *gin.Context, info *relaycommon.RelayIn
 }
 
 func (a *Adaptor) ConvertClaudeRequest(*gin.Context, *relaycommon.RelayInfo, *dto.ClaudeRequest) (any, error) {
-	return nil, errors.New("codex channel: endpoint not supported")
+	return nil, errors.New("codex channel: /v1/messages endpoint not supported")
 }
 
 func (a *Adaptor) ConvertAudioRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.AudioRequest) (io.Reader, error) {
@@ -41,15 +41,15 @@ func (a *Adaptor) Init(info *relaycommon.RelayInfo) {
 }
 
 func (a *Adaptor) ConvertOpenAIRequest(c *gin.Context, info *relaycommon.RelayInfo, request *dto.GeneralOpenAIRequest) (any, error) {
-	return nil, errors.New("codex channel: endpoint not supported")
+	return nil, errors.New("codex channel: /v1/chat/completions endpoint not supported")
 }
 
 func (a *Adaptor) ConvertRerankRequest(c *gin.Context, relayMode int, request dto.RerankRequest) (any, error) {
-	return nil, errors.New("codex channel: endpoint not supported")
+	return nil, errors.New("codex channel: /v1/rerank endpoint not supported")
 }
 
 func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.EmbeddingRequest) (any, error) {
-	return nil, errors.New("codex channel: endpoint not supported")
+	return nil, errors.New("codex channel: /v1/embeddings endpoint not supported")
 }
 
 func (a *Adaptor) ConvertOpenAIResponsesRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.OpenAIResponsesRequest) (any, error) {

relay/channel/deepseek/adaptor.go

@@ -95,11 +95,8 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayFormat {
 	case types.RelayFormatClaude:
-		if info.IsStream {
-			return claude.ClaudeStreamHandler(c, resp, info)
-		} else {
-			return claude.ClaudeHandler(c, resp, info)
-		}
+		adaptor := claude.Adaptor{}
+		return adaptor.DoResponse(c, resp, info)
 	default:
 		adaptor := openai.Adaptor{}
 		return adaptor.DoResponse(c, resp, info)

relay/channel/moonshot/adaptor.go

@@ -102,11 +102,8 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayFormat {
 	case types.RelayFormatClaude:
-		if info.IsStream {
-			return claude.ClaudeStreamHandler(c, resp, info)
-		} else {
-			return claude.ClaudeHandler(c, resp, info)
-		}
+		adaptor := claude.Adaptor{}
+		return adaptor.DoResponse(c, resp, info)
 	default:
 		adaptor := openai.Adaptor{}
 		return adaptor.DoResponse(c, resp, info)

relay/channel/openai/adaptor.go

@@ -171,7 +171,9 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		url = strings.Replace(url, "{model}", info.UpstreamModelName, -1)
 		return url, nil
 	default:
-		if info.RelayFormat == types.RelayFormatClaude || info.RelayFormat == types.RelayFormatGemini {
+		if (info.RelayFormat == types.RelayFormatClaude || info.RelayFormat == types.RelayFormatGemini) &&
+			info.RelayMode != relayconstant.RelayModeResponses &&
+			info.RelayMode != relayconstant.RelayModeResponsesCompact {
 			return fmt.Sprintf("%s/v1/chat/completions", info.ChannelBaseUrl), nil
 		}
 		return relaycommon.GetFullRequestURL(info.ChannelBaseUrl, info.RequestURLPath, info.ChannelType), nil

relay/channel/openai/chat_via_responses.go

@@ -71,12 +71,22 @@ func OaiResponsesToChatHandler(c *gin.Context, info *relaycommon.RelayInfo, resp
 		chatResp.Usage = *usage
 	}
 
-	chatBody, err := common.Marshal(chatResp)
+	var responseBody []byte
+	switch info.RelayFormat {
+	case types.RelayFormatClaude:
+		claudeResp := service.ResponseOpenAI2Claude(chatResp, info)
+		responseBody, err = common.Marshal(claudeResp)
+	case types.RelayFormatGemini:
+		geminiResp := service.ResponseOpenAI2Gemini(chatResp, info)
+		responseBody, err = common.Marshal(geminiResp)
+	default:
+		responseBody, err = common.Marshal(chatResp)
+	}
 	if err != nil {
 		return nil, types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
 	}
 
-	service.IOCopyBytesGracefully(c, resp, chatBody)
+	service.IOCopyBytesGracefully(c, resp, responseBody)
 	return usage, nil
 }
 
@@ -106,14 +116,43 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 	toolCallArgsByID := make(map[string]string)
 	toolCallNameSent := make(map[string]bool)
 	toolCallCanonicalIDByItemID := make(map[string]string)
+	hasSentReasoningSummary := false
+	needsReasoningSummarySeparator := false
 	//reasoningSummaryTextByKey := make(map[string]string)
 
+	if info.RelayFormat == types.RelayFormatClaude && info.ClaudeConvertInfo == nil {
+		info.ClaudeConvertInfo = &relaycommon.ClaudeConvertInfo{LastMessagesType: relaycommon.LastMessageTypeNone}
+	}
+
+	sendChatChunk := func(chunk *dto.ChatCompletionsStreamResponse) bool {
+		if chunk == nil {
+			return true
+		}
+		if info.RelayFormat == types.RelayFormatOpenAI {
+			if err := helper.ObjectData(c, chunk); err != nil {
+				streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+				return false
+			}
+			return true
+		}
+
+		chunkData, err := common.Marshal(chunk)
+		if err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
+			return false
+		}
+		if err := HandleStreamFormat(c, info, string(chunkData), false, false); err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+			return false
+		}
+		return true
+	}
+
 	sendStartIfNeeded := func() bool {
 		if sentStart {
 			return true
 		}
-		if err := helper.ObjectData(c, helper.GenerateStartEmptyResponse(responseId, createAt, model, nil)); err != nil {
-			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+		if !sendChatChunk(helper.GenerateStartEmptyResponse(responseId, createAt, model, nil)) {
 			return false
 		}
 		sentStart = true
@@ -154,6 +193,17 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 		if delta == "" {
 			return true
 		}
+		if needsReasoningSummarySeparator {
+			if strings.HasPrefix(delta, "\n\n") {
+				needsReasoningSummarySeparator = false
+			} else if strings.HasPrefix(delta, "\n") {
+				delta = "\n" + delta
+				needsReasoningSummarySeparator = false
+			} else {
+				delta = "\n\n" + delta
+				needsReasoningSummarySeparator = false
+			}
+		}
 		if !sendStartIfNeeded() {
 			return false
 		}
@@ -173,10 +223,10 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 				},
 			},
 		}
-		if err := helper.ObjectData(c, chunk); err != nil {
-			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+		if !sendChatChunk(chunk) {
 			return false
 		}
+		hasSentReasoningSummary = true
 		return true
 	}
 
@@ -231,8 +281,7 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 				},
 			},
 		}
-		if err := helper.ObjectData(c, chunk); err != nil {
-			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+		if !sendChatChunk(chunk) {
 			return false
 		}
 		sawToolCall = true
@@ -282,6 +331,9 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 			}
 
 		case "response.reasoning_summary_text.done":
+			if hasSentReasoningSummary {
+				needsReasoningSummarySeparator = true
+			}
 
 		//case "response.reasoning_summary_part.added", "response.reasoning_summary_part.done":
 		//	key := responsesStreamIndexKey(strings.TrimSpace(streamResp.ItemID), streamResp.SummaryIndex)
@@ -323,8 +375,7 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 						},
 					},
 				}
-				if err := helper.ObjectData(c, chunk); err != nil {
-					streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+				if !sendChatChunk(chunk) {
 					return false
 				}
 			}
@@ -419,13 +470,15 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 				return false
 			}
 			if !sentStop {
+				if info.RelayFormat == types.RelayFormatClaude && info.ClaudeConvertInfo != nil {
+					info.ClaudeConvertInfo.Usage = usage
+				}
 				finishReason := "stop"
 				if sawToolCall && outputText.Len() == 0 {
 					finishReason = "tool_calls"
 				}
 				stop := helper.GenerateStopResponse(responseId, createAt, model, finishReason)
-				if err := helper.ObjectData(c, stop); err != nil {
-					streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+				if !sendChatChunk(stop) {
 					return false
 				}
 				sentStop = true
@@ -456,26 +509,31 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 	}
 
 	if !sentStart {
-		if err := helper.ObjectData(c, helper.GenerateStartEmptyResponse(responseId, createAt, model, nil)); err != nil {
-			return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+		if !sendChatChunk(helper.GenerateStartEmptyResponse(responseId, createAt, model, nil)) {
+			return nil, streamErr
 		}
 	}
 	if !sentStop {
+		if info.RelayFormat == types.RelayFormatClaude && info.ClaudeConvertInfo != nil {
+			info.ClaudeConvertInfo.Usage = usage
+		}
 		finishReason := "stop"
 		if sawToolCall && outputText.Len() == 0 {
 			finishReason = "tool_calls"
 		}
 		stop := helper.GenerateStopResponse(responseId, createAt, model, finishReason)
-		if err := helper.ObjectData(c, stop); err != nil {
-			return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+		if !sendChatChunk(stop) {
+			return nil, streamErr
 		}
 	}
-	if info.ShouldIncludeUsage && usage != nil {
+	if info.RelayFormat == types.RelayFormatOpenAI && info.ShouldIncludeUsage && usage != nil {
 		if err := helper.ObjectData(c, helper.GenerateFinalUsageResponse(responseId, createAt, model, *usage)); err != nil {
 			return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
 		}
 	}
 
-	helper.Done(c)
+	if info.RelayFormat == types.RelayFormatOpenAI {
+		helper.Done(c)
+	}
 	return usage, nil
 }

relay/channel/vertex/adaptor.go

@@ -365,10 +365,11 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 }
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
+	claudeAdaptor := claude.Adaptor{}
 	if info.IsStream {
 		switch a.RequestMode {
 		case RequestModeClaude:
-			return claude.ClaudeStreamHandler(c, resp, info)
+			return claudeAdaptor.DoResponse(c, resp, info)
 		case RequestModeGemini:
 			if info.RelayMode == constant.RelayModeGemini {
 				return gemini.GeminiTextGenerationStreamHandler(c, info, resp)
@@ -381,7 +382,7 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 	} else {
 		switch a.RequestMode {
 		case RequestModeClaude:
-			return claude.ClaudeHandler(c, resp, info)
+			return claudeAdaptor.DoResponse(c, resp, info)
 		case RequestModeGemini:
 			if info.RelayMode == constant.RelayModeGemini {
 				return gemini.GeminiTextGenerationHandler(c, info, resp)

relay/channel/volcengine/adaptor.go

@@ -347,10 +347,8 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	if info.RelayFormat == types.RelayFormatClaude {
 		if _, ok := channelconstant.ChannelSpecialBases[info.ChannelBaseUrl]; ok {
-			if info.IsStream {
-				return claude.ClaudeStreamHandler(c, resp, info)
-			}
-			return claude.ClaudeHandler(c, resp, info)
+			adaptor := claude.Adaptor{}
+			return adaptor.DoResponse(c, resp, info)
 		}
 	}
 

relay/channel/zhipu_4v/adaptor.go

@@ -109,11 +109,8 @@ func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, request
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *types.NewAPIError) {
 	switch info.RelayFormat {
 	case types.RelayFormatClaude:
-		if info.IsStream {
-			return claude.ClaudeStreamHandler(c, resp, info)
-		} else {
-			return claude.ClaudeHandler(c, resp, info)
-		}
+		adaptor := claude.Adaptor{}
+		return adaptor.DoResponse(c, resp, info)
 	default:
 		if info.RelayMode == relayconstant.RelayModeImagesGenerations {
 			return zhipu4vImageHandler(c, resp, info)

relay/claude_handler.go

@@ -110,6 +110,23 @@ func ClaudeHelper(c *gin.Context, info *relaycommon.RelayInfo) (newAPIError *typ
 		}
 	}
 
+	if !model_setting.GetGlobalSettings().PassThroughRequestEnabled &&
+		!info.ChannelSetting.PassThroughBodyEnabled &&
+		service.ShouldChatCompletionsUseResponsesGlobal(info.ChannelId, info.ChannelType, info.OriginModelName) {
+		openAIRequest, convErr := service.ClaudeToOpenAIRequest(*request, info)
+		if convErr != nil {
+			return types.NewError(convErr, types.ErrorCodeConvertRequestFailed, types.ErrOptionWithSkipRetry())
+		}
+
+		usage, newApiErr := chatCompletionsViaResponses(c, info, adaptor, openAIRequest)
+		if newApiErr != nil {
+			return newApiErr
+		}
+
+		service.PostClaudeConsumeQuota(c, info, usage)
+		return nil
+	}
+
 	var requestBody io.Reader
 	if model_setting.GetGlobalSettings().PassThroughRequestEnabled || info.ChannelSetting.PassThroughBodyEnabled {
 		body, err := common.GetRequestBody(c)

relay/common/relay_info.go

@@ -37,6 +37,9 @@ type ClaudeConvertInfo struct {
 	Usage            *dto.Usage
 	FinishReason     string
 	Done             bool
+
+	ToolCallBaseIndex      int
+	ToolCallMaxIndexOffset int
 }
 
 type RerankerInfo struct {
@@ -145,6 +148,8 @@ type RelayInfo struct {
 	// RequestConversionChain records request format conversions in order, e.g.
 	// ["openai", "openai_responses"] or ["openai", "claude"].
 	RequestConversionChain []types.RelayFormat
+	// 最终请求到上游的格式 TODO: 当前仅设置了Claude
+	FinalRequestRelayFormat types.RelayFormat
 
 	ThinkingContentInfo
 	TokenCountMeta
@@ -319,12 +324,15 @@ func GenRelayInfoClaude(c *gin.Context, request dto.Request) *RelayInfo {
 	info.ClaudeConvertInfo = &ClaudeConvertInfo{
 		LastMessagesType: LastMessageTypeNone,
 	}
-	if c.Query("beta") == "true" {
-		info.IsClaudeBetaQuery = true
-	}
+	info.IsClaudeBetaQuery = c.Query("beta") == "true" || isClaudeBetaForced(c)
 	return info
 }
 
+func isClaudeBetaForced(c *gin.Context) bool {
+	channelOtherSettings, ok := common.GetContextKeyType[dto.ChannelOtherSettings](c, constant.ContextKeyChannelOtherSetting)
+	return ok && channelOtherSettings.ClaudeBetaQuery
+}
+
 func GenRelayInfoRerank(c *gin.Context, request *dto.RerankRequest) *RelayInfo {
 	info := genBaseRelayInfo(c, request)
 	info.RelayMode = relayconstant.RelayModeRerank

relay/compatible_handler.go

@@ -334,7 +334,7 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, usage
 
 	var audioInputQuota decimal.Decimal
 	var audioInputPrice float64
-	isClaudeUsageSemantic := relayInfo.ChannelType == constant.ChannelTypeAnthropic
+	isClaudeUsageSemantic := relayInfo.FinalRequestRelayFormat == types.RelayFormatClaude
 	if !relayInfo.PriceData.UsePrice {
 		baseTokens := dPromptTokens
 		// 减去 cached tokens

service/convert.go

@@ -207,6 +207,44 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 	}
 
 	var claudeResponses []*dto.ClaudeResponse
+	// stopOpenBlocks emits the required content_block_stop event(s) for the currently open block(s)
+	// according to Anthropic's SSE streaming state machine:
+	// content_block_start -> content_block_delta* -> content_block_stop (per index).
+	//
+	// For text/thinking, there is at most one open block at info.ClaudeConvertInfo.Index.
+	// For tools, OpenAI tool_calls can stream multiple parallel tool_use blocks (indexed from 0),
+	// so we may have multiple open blocks and must stop each one explicitly.
+	stopOpenBlocks := func() {
+		switch info.ClaudeConvertInfo.LastMessagesType {
+		case relaycommon.LastMessageTypeText, relaycommon.LastMessageTypeThinking:
+			claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
+		case relaycommon.LastMessageTypeTools:
+			base := info.ClaudeConvertInfo.ToolCallBaseIndex
+			for offset := 0; offset <= info.ClaudeConvertInfo.ToolCallMaxIndexOffset; offset++ {
+				claudeResponses = append(claudeResponses, generateStopBlock(base+offset))
+			}
+		}
+	}
+	// stopOpenBlocksAndAdvance closes the currently open block(s) and advances the content block index
+	// to the next available slot for subsequent content_block_start events.
+	//
+	// This prevents invalid streams where a content_block_delta (e.g. thinking_delta) is emitted for an
+	// index whose active content_block type is different (the typical cause of "Mismatched content block type").
+	stopOpenBlocksAndAdvance := func() {
+		if info.ClaudeConvertInfo.LastMessagesType == relaycommon.LastMessageTypeNone {
+			return
+		}
+		stopOpenBlocks()
+		switch info.ClaudeConvertInfo.LastMessagesType {
+		case relaycommon.LastMessageTypeTools:
+			info.ClaudeConvertInfo.Index = info.ClaudeConvertInfo.ToolCallBaseIndex + info.ClaudeConvertInfo.ToolCallMaxIndexOffset + 1
+			info.ClaudeConvertInfo.ToolCallBaseIndex = 0
+			info.ClaudeConvertInfo.ToolCallMaxIndexOffset = 0
+		default:
+			info.ClaudeConvertInfo.Index++
+		}
+		info.ClaudeConvertInfo.LastMessagesType = relaycommon.LastMessageTypeNone
+	}
 	if info.SendResponseCount == 1 {
 		msg := &dto.ClaudeMediaMessage{
 			Id:    openAIResponse.Id,
@@ -228,6 +266,8 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 		//})
 		if openAIResponse.IsToolCall() {
 			info.ClaudeConvertInfo.LastMessagesType = relaycommon.LastMessageTypeTools
+			info.ClaudeConvertInfo.ToolCallBaseIndex = 0
+			info.ClaudeConvertInfo.ToolCallMaxIndexOffset = 0
 			var toolCall dto.ToolCallResponse
 			if len(openAIResponse.Choices) > 0 && len(openAIResponse.Choices[0].Delta.ToolCalls) > 0 {
 				toolCall = openAIResponse.Choices[0].Delta.ToolCalls[0]
@@ -252,8 +292,9 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 			claudeResponses = append(claudeResponses, resp)
 			// 首块包含工具 delta，则追加 input_json_delta
 			if toolCall.Function.Arguments != "" {
+				idx := 0
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-					Index: &info.ClaudeConvertInfo.Index,
+					Index: &idx,
 					Type:  "content_block_delta",
 					Delta: &dto.ClaudeMediaMessage{
 						Type:        "input_json_delta",
@@ -270,16 +311,21 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 			content := openAIResponse.Choices[0].Delta.GetContentString()
 
 			if reasoning != "" {
+				if info.ClaudeConvertInfo.LastMessagesType != relaycommon.LastMessageTypeThinking {
+					stopOpenBlocksAndAdvance()
+				}
+				idx := info.ClaudeConvertInfo.Index
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-					Index: &info.ClaudeConvertInfo.Index,
+					Index: &idx,
 					Type:  "content_block_start",
 					ContentBlock: &dto.ClaudeMediaMessage{
 						Type:     "thinking",
 						Thinking: common.GetPointer[string](""),
 					},
 				})
+				idx2 := idx
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-					Index: &info.ClaudeConvertInfo.Index,
+					Index: &idx2,
 					Type:  "content_block_delta",
 					Delta: &dto.ClaudeMediaMessage{
 						Type:     "thinking_delta",
@@ -288,16 +334,21 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 				})
 				info.ClaudeConvertInfo.LastMessagesType = relaycommon.LastMessageTypeThinking
 			} else if content != "" {
+				if info.ClaudeConvertInfo.LastMessagesType != relaycommon.LastMessageTypeText {
+					stopOpenBlocksAndAdvance()
+				}
+				idx := info.ClaudeConvertInfo.Index
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-					Index: &info.ClaudeConvertInfo.Index,
+					Index: &idx,
 					Type:  "content_block_start",
 					ContentBlock: &dto.ClaudeMediaMessage{
 						Type: "text",
 						Text: common.GetPointer[string](""),
 					},
 				})
+				idx2 := idx
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-					Index: &info.ClaudeConvertInfo.Index,
+					Index: &idx2,
 					Type:  "content_block_delta",
 					Delta: &dto.ClaudeMediaMessage{
 						Type: "text_delta",
@@ -311,7 +362,7 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 		// 如果首块就带 finish_reason，需要立即发送停止块
 		if len(openAIResponse.Choices) > 0 && openAIResponse.Choices[0].FinishReason != nil && *openAIResponse.Choices[0].FinishReason != "" {
 			info.FinishReason = *openAIResponse.Choices[0].FinishReason
-			claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
+			stopOpenBlocks()
 			oaiUsage := openAIResponse.Usage
 			if oaiUsage == nil {
 				oaiUsage = info.ClaudeConvertInfo.Usage
@@ -342,7 +393,7 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 		// no choices
 		// 可能为非标准的 OpenAI 响应，判断是否已经完成
 		if info.ClaudeConvertInfo.Done {
-			claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
+			stopOpenBlocks()
 			oaiUsage := info.ClaudeConvertInfo.Usage
 			if oaiUsage != nil {
 				claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
@@ -376,18 +427,25 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 		if len(chosenChoice.Delta.ToolCalls) > 0 {
 			toolCalls := chosenChoice.Delta.ToolCalls
 			if info.ClaudeConvertInfo.LastMessagesType != relaycommon.LastMessageTypeTools {
-				claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
-				info.ClaudeConvertInfo.Index++
+				stopOpenBlocksAndAdvance()
+				info.ClaudeConvertInfo.ToolCallBaseIndex = info.ClaudeConvertInfo.Index
+				info.ClaudeConvertInfo.ToolCallMaxIndexOffset = 0
 			}
 			info.ClaudeConvertInfo.LastMessagesType = relaycommon.LastMessageTypeTools
+			base := info.ClaudeConvertInfo.ToolCallBaseIndex
+			maxOffset := info.ClaudeConvertInfo.ToolCallMaxIndexOffset
 
 			for i, toolCall := range toolCalls {
-				blockIndex := info.ClaudeConvertInfo.Index
+				offset := 0
 				if toolCall.Index != nil {
-					blockIndex = *toolCall.Index
-				} else if len(toolCalls) > 1 {
-					blockIndex = info.ClaudeConvertInfo.Index + i
+					offset = *toolCall.Index
+				} else {
+					offset = i
 				}
+				if offset > maxOffset {
+					maxOffset = offset
+				}
+				blockIndex := base + offset
 
 				idx := blockIndex
 				if toolCall.Function.Name != "" {
@@ -413,17 +471,19 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 						},
 					})
 				}
-
-				info.ClaudeConvertInfo.Index = blockIndex
 			}
+			info.ClaudeConvertInfo.ToolCallMaxIndexOffset = maxOffset
+			info.ClaudeConvertInfo.Index = base + maxOffset
 		} else {
 			reasoning := chosenChoice.Delta.GetReasoningContent()
 			textContent := chosenChoice.Delta.GetContentString()
 			if reasoning != "" || textContent != "" {
 				if reasoning != "" {
 					if info.ClaudeConvertInfo.LastMessagesType != relaycommon.LastMessageTypeThinking {
+						stopOpenBlocksAndAdvance()
+						idx := info.ClaudeConvertInfo.Index
 						claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-							Index: &info.ClaudeConvertInfo.Index,
+							Index: &idx,
 							Type:  "content_block_start",
 							ContentBlock: &dto.ClaudeMediaMessage{
 								Type:     "thinking",
@@ -438,12 +498,10 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 					}
 				} else {
 					if info.ClaudeConvertInfo.LastMessagesType != relaycommon.LastMessageTypeText {
-						if info.ClaudeConvertInfo.LastMessagesType == relaycommon.LastMessageTypeThinking || info.ClaudeConvertInfo.LastMessagesType == relaycommon.LastMessageTypeTools {
-							claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
-							info.ClaudeConvertInfo.Index++
-						}
+						stopOpenBlocksAndAdvance()
+						idx := info.ClaudeConvertInfo.Index
 						claudeResponses = append(claudeResponses, &dto.ClaudeResponse{
-							Index: &info.ClaudeConvertInfo.Index,
+							Index: &idx,
 							Type:  "content_block_start",
 							ContentBlock: &dto.ClaudeMediaMessage{
 								Type: "text",
@@ -462,13 +520,13 @@ func StreamResponseOpenAI2Claude(openAIResponse *dto.ChatCompletionsStreamRespon
 			}
 		}
 
-		claudeResponse.Index = &info.ClaudeConvertInfo.Index
+		claudeResponse.Index = common.GetPointer[int](info.ClaudeConvertInfo.Index)
 		if !isEmpty && claudeResponse.Delta != nil {
 			claudeResponses = append(claudeResponses, &claudeResponse)
 		}
 
 		if doneChunk || info.ClaudeConvertInfo.Done {
-			claudeResponses = append(claudeResponses, generateStopBlock(info.ClaudeConvertInfo.Index))
+			stopOpenBlocks()
 			oaiUsage := openAIResponse.Usage
 			if oaiUsage == nil {
 				oaiUsage = info.ClaudeConvertInfo.Usage

setting/ratio_setting/cache_ratio.go

@@ -1,10 +1,7 @@
 package ratio_setting
 
 import (
-	"encoding/json"
-	"sync"
-
-	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/types"
 )
 
 var defaultCacheRatio = map[string]float64{
@@ -98,44 +95,37 @@ var defaultCreateCacheRatio = map[string]float64{
 
 //var defaultCreateCacheRatio = map[string]float64{}
 
-var cacheRatioMap map[string]float64
-var cacheRatioMapMutex sync.RWMutex
+var cacheRatioMap = types.NewRWMap[string, float64]()
+var createCacheRatioMap = types.NewRWMap[string, float64]()
 
-// GetCacheRatioMap returns the cache ratio map
+// GetCacheRatioMap returns a copy of the cache ratio map
 func GetCacheRatioMap() map[string]float64 {
-	cacheRatioMapMutex.RLock()
-	defer cacheRatioMapMutex.RUnlock()
-	return cacheRatioMap
+	return cacheRatioMap.ReadAll()
 }
 
 // CacheRatio2JSONString converts the cache ratio map to a JSON string
 func CacheRatio2JSONString() string {
-	cacheRatioMapMutex.RLock()
-	defer cacheRatioMapMutex.RUnlock()
-	jsonBytes, err := json.Marshal(cacheRatioMap)
-	if err != nil {
-		common.SysLog("error marshalling cache ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return cacheRatioMap.MarshalJSONString()
+}
+
+// CreateCacheRatio2JSONString converts the create cache ratio map to a JSON string
+func CreateCacheRatio2JSONString() string {
+	return createCacheRatioMap.MarshalJSONString()
 }
 
 // UpdateCacheRatioByJSONString updates the cache ratio map from a JSON string
 func UpdateCacheRatioByJSONString(jsonStr string) error {
-	cacheRatioMapMutex.Lock()
-	defer cacheRatioMapMutex.Unlock()
-	cacheRatioMap = make(map[string]float64)
-	err := json.Unmarshal([]byte(jsonStr), &cacheRatioMap)
-	if err == nil {
-		InvalidateExposedDataCache()
-	}
-	return err
+	return types.LoadFromJsonStringWithCallback(cacheRatioMap, jsonStr, InvalidateExposedDataCache)
+}
+
+// UpdateCreateCacheRatioByJSONString updates the create cache ratio map from a JSON string
+func UpdateCreateCacheRatioByJSONString(jsonStr string) error {
+	return types.LoadFromJsonStringWithCallback(createCacheRatioMap, jsonStr, InvalidateExposedDataCache)
 }
 
 // GetCacheRatio returns the cache ratio for a model
 func GetCacheRatio(name string) (float64, bool) {
-	cacheRatioMapMutex.RLock()
-	defer cacheRatioMapMutex.RUnlock()
-	ratio, ok := cacheRatioMap[name]
+	ratio, ok := cacheRatioMap.Get(name)
 	if !ok {
 		return 1, false // Default to 1 if not found
 	}
@@ -143,7 +133,7 @@ func GetCacheRatio(name string) (float64, bool) {
 }
 
 func GetCreateCacheRatio(name string) (float64, bool) {
-	ratio, ok := defaultCreateCacheRatio[name]
+	ratio, ok := createCacheRatioMap.Get(name)
 	if !ok {
 		return 1.25, false // Default to 1.25 if not found
 	}
@@ -151,11 +141,9 @@ func GetCreateCacheRatio(name string) (float64, bool) {
 }
 
 func GetCacheRatioCopy() map[string]float64 {
-	cacheRatioMapMutex.RLock()
-	defer cacheRatioMapMutex.RUnlock()
-	copyMap := make(map[string]float64, len(cacheRatioMap))
-	for k, v := range cacheRatioMap {
-		copyMap[k] = v
-	}
-	return copyMap
+	return cacheRatioMap.ReadAll()
+}
+
+func GetCreateCacheRatioCopy() map[string]float64 {
+	return createCacheRatioMap.ReadAll()
 }

setting/ratio_setting/exposed_cache.go

@@ -42,10 +42,11 @@ func GetExposedData() gin.H {
 		return cloneGinH(c.data)
 	}
 	newData := gin.H{
-		"model_ratio":      GetModelRatioCopy(),
-		"completion_ratio": GetCompletionRatioCopy(),
-		"cache_ratio":      GetCacheRatioCopy(),
-		"model_price":      GetModelPriceCopy(),
+		"model_ratio":        GetModelRatioCopy(),
+		"completion_ratio":   GetCompletionRatioCopy(),
+		"cache_ratio":        GetCacheRatioCopy(),
+		"create_cache_ratio": GetCreateCacheRatioCopy(),
+		"model_price":        GetModelPriceCopy(),
 	}
 	exposedData.Store(&exposedCache{
 		data:      newData,

setting/ratio_setting/group_ratio.go

@@ -3,29 +3,27 @@ package ratio_setting
 import (
 	"encoding/json"
 	"errors"
-	"sync"
 
 	"github.com/QuantumNous/new-api/common"
 	"github.com/QuantumNous/new-api/setting/config"
 	"github.com/QuantumNous/new-api/types"
 )
 
-var groupRatio = map[string]float64{
+var defaultGroupRatio = map[string]float64{
 	"default": 1,
 	"vip":     1,
 	"svip":    1,
 }
 
-var groupRatioMutex sync.RWMutex
+var groupRatioMap = types.NewRWMap[string, float64]()
 
-var (
-	GroupGroupRatio = map[string]map[string]float64{
-		"vip": {
-			"edit_this": 0.9,
-		},
-	}
-	groupGroupRatioMutex sync.RWMutex
-)
+var defaultGroupGroupRatio = map[string]map[string]float64{
+	"vip": {
+		"edit_this": 0.9,
+	},
+}
+
+var groupGroupRatioMap = types.NewRWMap[string, map[string]float64]()
 
 var defaultGroupSpecialUsableGroup = map[string]map[string]string{
 	"vip": {
@@ -35,9 +33,9 @@ var defaultGroupSpecialUsableGroup = map[string]map[string]string{
 }
 
 type GroupRatioSetting struct {
-	GroupRatio              map[string]float64                      `json:"group_ratio"`
-	GroupGroupRatio         map[string]map[string]float64           `json:"group_group_ratio"`
-	GroupSpecialUsableGroup *types.RWMap[string, map[string]string] `json:"group_special_usable_group"`
+	GroupRatio              *types.RWMap[string, float64]            `json:"group_ratio"`
+	GroupGroupRatio         *types.RWMap[string, map[string]float64] `json:"group_group_ratio"`
+	GroupSpecialUsableGroup *types.RWMap[string, map[string]string]  `json:"group_special_usable_group"`
 }
 
 var groupRatioSetting GroupRatioSetting
@@ -46,10 +44,13 @@ func init() {
 	groupSpecialUsableGroup := types.NewRWMap[string, map[string]string]()
 	groupSpecialUsableGroup.AddAll(defaultGroupSpecialUsableGroup)
 
+	groupRatioMap.AddAll(defaultGroupRatio)
+	groupGroupRatioMap.AddAll(defaultGroupGroupRatio)
+
 	groupRatioSetting = GroupRatioSetting{
 		GroupSpecialUsableGroup: groupSpecialUsableGroup,
-		GroupRatio:              groupRatio,
-		GroupGroupRatio:         GroupGroupRatio,
+		GroupRatio:              groupRatioMap,
+		GroupGroupRatio:         groupGroupRatioMap,
 	}
 
 	config.GlobalConfig.Register("group_ratio_setting", &groupRatioSetting)
@@ -64,48 +65,24 @@ func GetGroupRatioSetting() *GroupRatioSetting {
 }
 
 func GetGroupRatioCopy() map[string]float64 {
-	groupRatioMutex.RLock()
-	defer groupRatioMutex.RUnlock()
-
-	groupRatioCopy := make(map[string]float64)
-	for k, v := range groupRatio {
-		groupRatioCopy[k] = v
-	}
-	return groupRatioCopy
+	return groupRatioMap.ReadAll()
 }
 
 func ContainsGroupRatio(name string) bool {
-	groupRatioMutex.RLock()
-	defer groupRatioMutex.RUnlock()
-
-	_, ok := groupRatio[name]
+	_, ok := groupRatioMap.Get(name)
 	return ok
 }
 
 func GroupRatio2JSONString() string {
-	groupRatioMutex.RLock()
-	defer groupRatioMutex.RUnlock()
-
-	jsonBytes, err := json.Marshal(groupRatio)
-	if err != nil {
-		common.SysLog("error marshalling model ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return groupRatioMap.MarshalJSONString()
 }
 
 func UpdateGroupRatioByJSONString(jsonStr string) error {
-	groupRatioMutex.Lock()
-	defer groupRatioMutex.Unlock()
-
-	groupRatio = make(map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &groupRatio)
+	return types.LoadFromJsonString(groupRatioMap, jsonStr)
 }
 
 func GetGroupRatio(name string) float64 {
-	groupRatioMutex.RLock()
-	defer groupRatioMutex.RUnlock()
-
-	ratio, ok := groupRatio[name]
+	ratio, ok := groupRatioMap.Get(name)
 	if !ok {
 		common.SysLog("group ratio not found: " + name)
 		return 1
@@ -114,10 +91,7 @@ func GetGroupRatio(name string) float64 {
 }
 
 func GetGroupGroupRatio(userGroup, usingGroup string) (float64, bool) {
-	groupGroupRatioMutex.RLock()
-	defer groupGroupRatioMutex.RUnlock()
-
-	gp, ok := GroupGroupRatio[userGroup]
+	gp, ok := groupGroupRatioMap.Get(userGroup)
 	if !ok {
 		return -1, false
 	}
@@ -129,22 +103,11 @@ func GetGroupGroupRatio(userGroup, usingGroup string) (float64, bool) {
 }
 
 func GroupGroupRatio2JSONString() string {
-	groupGroupRatioMutex.RLock()
-	defer groupGroupRatioMutex.RUnlock()
-
-	jsonBytes, err := json.Marshal(GroupGroupRatio)
-	if err != nil {
-		common.SysLog("error marshalling group-group ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return groupGroupRatioMap.MarshalJSONString()
 }
 
 func UpdateGroupGroupRatioByJSONString(jsonStr string) error {
-	groupGroupRatioMutex.Lock()
-	defer groupGroupRatioMutex.Unlock()
-
-	GroupGroupRatio = make(map[string]map[string]float64)
-	return json.Unmarshal([]byte(jsonStr), &GroupGroupRatio)
+	return types.LoadFromJsonString(groupGroupRatioMap, jsonStr)
 }
 
 func CheckGroupRatio(jsonStr string) error {

setting/ratio_setting/model_ratio.go

@@ -1,12 +1,11 @@
 package ratio_setting
 
 import (
-	"encoding/json"
 	"strings"
-	"sync"
 
 	"github.com/QuantumNous/new-api/common"
 	"github.com/QuantumNous/new-api/setting/operation_setting"
+	"github.com/QuantumNous/new-api/types"
 )
 
 // from songquanpeng/one-api
@@ -319,19 +318,9 @@ var defaultAudioCompletionRatio = map[string]float64{
 	"tts-1-hd-1106":        0,
 }
 
-var (
-	modelPriceMap      map[string]float64 = nil
-	modelPriceMapMutex                    = sync.RWMutex{}
-)
-var (
-	modelRatioMap      map[string]float64 = nil
-	modelRatioMapMutex                    = sync.RWMutex{}
-)
-
-var (
-	CompletionRatio      map[string]float64 = nil
-	CompletionRatioMutex                    = sync.RWMutex{}
-)
+var modelPriceMap = types.NewRWMap[string, float64]()
+var modelRatioMap = types.NewRWMap[string, float64]()
+var completionRatioMap = types.NewRWMap[string, float64]()
 
 var defaultCompletionRatio = map[string]float64{
 	"gpt-4-gizmo-*":  2,
@@ -342,79 +331,34 @@ var defaultCompletionRatio = map[string]float64{
 
 // InitRatioSettings initializes all model related settings maps
 func InitRatioSettings() {
-	// Initialize modelPriceMap
-	modelPriceMapMutex.Lock()
-	modelPriceMap = defaultModelPrice
-	modelPriceMapMutex.Unlock()
-
-	// Initialize modelRatioMap
-	modelRatioMapMutex.Lock()
-	modelRatioMap = defaultModelRatio
-	modelRatioMapMutex.Unlock()
-
-	// Initialize CompletionRatio
-	CompletionRatioMutex.Lock()
-	CompletionRatio = defaultCompletionRatio
-	CompletionRatioMutex.Unlock()
-
-	// Initialize cacheRatioMap
-	cacheRatioMapMutex.Lock()
-	cacheRatioMap = defaultCacheRatio
-	cacheRatioMapMutex.Unlock()
-
-	// initialize imageRatioMap
-	imageRatioMapMutex.Lock()
-	imageRatioMap = defaultImageRatio
-	imageRatioMapMutex.Unlock()
-
-	// initialize audioRatioMap
-	audioRatioMapMutex.Lock()
-	audioRatioMap = defaultAudioRatio
-	audioRatioMapMutex.Unlock()
-
-	// initialize audioCompletionRatioMap
-	audioCompletionRatioMapMutex.Lock()
-	audioCompletionRatioMap = defaultAudioCompletionRatio
-	audioCompletionRatioMapMutex.Unlock()
+	modelPriceMap.AddAll(defaultModelPrice)
+	modelRatioMap.AddAll(defaultModelRatio)
+	completionRatioMap.AddAll(defaultCompletionRatio)
+	cacheRatioMap.AddAll(defaultCacheRatio)
+	createCacheRatioMap.AddAll(defaultCreateCacheRatio)
+	imageRatioMap.AddAll(defaultImageRatio)
+	audioRatioMap.AddAll(defaultAudioRatio)
+	audioCompletionRatioMap.AddAll(defaultAudioCompletionRatio)
 }
 
 func GetModelPriceMap() map[string]float64 {
-	modelPriceMapMutex.RLock()
-	defer modelPriceMapMutex.RUnlock()
-	return modelPriceMap
+	return modelPriceMap.ReadAll()
 }
 
 func ModelPrice2JSONString() string {
-	modelPriceMapMutex.RLock()
-	defer modelPriceMapMutex.RUnlock()
-
-	jsonBytes, err := common.Marshal(modelPriceMap)
-	if err != nil {
-		common.SysError("error marshalling model price: " + err.Error())
-	}
-	return string(jsonBytes)
+	return modelPriceMap.MarshalJSONString()
 }
 
 func UpdateModelPriceByJSONString(jsonStr string) error {
-	modelPriceMapMutex.Lock()
-	defer modelPriceMapMutex.Unlock()
-	modelPriceMap = make(map[string]float64)
-	err := json.Unmarshal([]byte(jsonStr), &modelPriceMap)
-	if err == nil {
-		InvalidateExposedDataCache()
-	}
-	return err
+	return types.LoadFromJsonStringWithCallback(modelPriceMap, jsonStr, InvalidateExposedDataCache)
 }
 
 // GetModelPrice 返回模型的价格，如果模型不存在则返回-1，false
 func GetModelPrice(name string, printErr bool) (float64, bool) {
-	modelPriceMapMutex.RLock()
-	defer modelPriceMapMutex.RUnlock()
-
 	name = FormatMatchingModelName(name)
 
 	if strings.HasSuffix(name, CompactModelSuffix) {
-		price, ok := modelPriceMap[CompactWildcardModelKey]
+		price, ok := modelPriceMap.Get(CompactWildcardModelKey)
 		if !ok {
 			if printErr {
 				common.SysError("model price not found: " + name)
@@ -424,7 +368,7 @@ func GetModelPrice(name string, printErr bool) (float64, bool) {
 		return price, true
 	}
 
-	price, ok := modelPriceMap[name]
+	price, ok := modelPriceMap.Get(name)
 	if !ok {
 		if printErr {
 			common.SysError("model price not found: " + name)
@@ -435,14 +379,7 @@ func GetModelPrice(name string, printErr bool) (float64, bool) {
 }
 
 func UpdateModelRatioByJSONString(jsonStr string) error {
-	modelRatioMapMutex.Lock()
-	defer modelRatioMapMutex.Unlock()
-	modelRatioMap = make(map[string]float64)
-	err := common.Unmarshal([]byte(jsonStr), &modelRatioMap)
-	if err == nil {
-		InvalidateExposedDataCache()
-	}
-	return err
+	return types.LoadFromJsonStringWithCallback(modelRatioMap, jsonStr, InvalidateExposedDataCache)
 }
 
 // 处理带有思考预算的模型名称，方便统一定价
@@ -454,15 +391,12 @@ func handleThinkingBudgetModel(name, prefix, wildcard string) string {
 }
 
 func GetModelRatio(name string) (float64, bool, string) {
-	modelRatioMapMutex.RLock()
-	defer modelRatioMapMutex.RUnlock()
-
 	name = FormatMatchingModelName(name)
 
-	ratio, ok := modelRatioMap[name]
+	ratio, ok := modelRatioMap.Get(name)
 	if !ok {
 		if strings.HasSuffix(name, CompactModelSuffix) {
-			if wildcardRatio, ok := modelRatioMap[CompactWildcardModelKey]; ok {
+			if wildcardRatio, ok := modelRatioMap.Get(CompactWildcardModelKey); ok {
 				return wildcardRatio, true, name
 			}
 			//return 0, true, name
@@ -488,54 +422,19 @@ func GetDefaultModelPriceMap() map[string]float64 {
 	return defaultModelPrice
 }
 
-func GetDefaultImageRatioMap() map[string]float64 {
-	return defaultImageRatio
-}
-
-func GetDefaultAudioRatioMap() map[string]float64 {
-	return defaultAudioRatio
-}
-
-func GetDefaultAudioCompletionRatioMap() map[string]float64 {
-	return defaultAudioCompletionRatio
-}
-
-func GetCompletionRatioMap() map[string]float64 {
-	CompletionRatioMutex.RLock()
-	defer CompletionRatioMutex.RUnlock()
-	return CompletionRatio
-}
-
 func CompletionRatio2JSONString() string {
-	CompletionRatioMutex.RLock()
-	defer CompletionRatioMutex.RUnlock()
-
-	jsonBytes, err := json.Marshal(CompletionRatio)
-	if err != nil {
-		common.SysError("error marshalling completion ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return completionRatioMap.MarshalJSONString()
 }
 
 func UpdateCompletionRatioByJSONString(jsonStr string) error {
-	CompletionRatioMutex.Lock()
-	defer CompletionRatioMutex.Unlock()
-	CompletionRatio = make(map[string]float64)
-	err := common.Unmarshal([]byte(jsonStr), &CompletionRatio)
-	if err == nil {
-		InvalidateExposedDataCache()
-	}
-	return err
+	return types.LoadFromJsonStringWithCallback(completionRatioMap, jsonStr, InvalidateExposedDataCache)
 }
 
 func GetCompletionRatio(name string) float64 {
-	CompletionRatioMutex.RLock()
-	defer CompletionRatioMutex.RUnlock()
-
 	name = FormatMatchingModelName(name)
 
 	if strings.Contains(name, "/") {
-		if ratio, ok := CompletionRatio[name]; ok {
+		if ratio, ok := completionRatioMap.Get(name); ok {
 			return ratio
 		}
 	}
@@ -543,7 +442,7 @@ func GetCompletionRatio(name string) float64 {
 	if contain {
 		return hardCodedRatio
 	}
-	if ratio, ok := CompletionRatio[name]; ok {
+	if ratio, ok := completionRatioMap.Get(name); ok {
 		return ratio
 	}
 	return hardCodedRatio
@@ -671,88 +570,54 @@ func getHardcodedCompletionModelRatio(name string) (float64, bool) {
 }
 
 func GetAudioRatio(name string) float64 {
-	audioRatioMapMutex.RLock()
-	defer audioRatioMapMutex.RUnlock()
 	name = FormatMatchingModelName(name)
-	if ratio, ok := audioRatioMap[name]; ok {
+	if ratio, ok := audioRatioMap.Get(name); ok {
 		return ratio
 	}
 	return 1
 }
 
 func GetAudioCompletionRatio(name string) float64 {
-	audioCompletionRatioMapMutex.RLock()
-	defer audioCompletionRatioMapMutex.RUnlock()
 	name = FormatMatchingModelName(name)
-	if ratio, ok := audioCompletionRatioMap[name]; ok {
-
+	if ratio, ok := audioCompletionRatioMap.Get(name); ok {
 		return ratio
 	}
 	return 1
 }
 
 func ContainsAudioRatio(name string) bool {
-	audioRatioMapMutex.RLock()
-	defer audioRatioMapMutex.RUnlock()
 	name = FormatMatchingModelName(name)
-	_, ok := audioRatioMap[name]
+	_, ok := audioRatioMap.Get(name)
 	return ok
 }
 
 func ContainsAudioCompletionRatio(name string) bool {
-	audioCompletionRatioMapMutex.RLock()
-	defer audioCompletionRatioMapMutex.RUnlock()
 	name = FormatMatchingModelName(name)
-	_, ok := audioCompletionRatioMap[name]
+	_, ok := audioCompletionRatioMap.Get(name)
 	return ok
 }
 
 func ModelRatio2JSONString() string {
-	modelRatioMapMutex.RLock()
-	defer modelRatioMapMutex.RUnlock()
-
-	jsonBytes, err := common.Marshal(modelRatioMap)
-	if err != nil {
-		common.SysError("error marshalling model ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return modelRatioMap.MarshalJSONString()
 }
 
 var defaultImageRatio = map[string]float64{
 	"gpt-image-1": 2,
 }
-var imageRatioMap map[string]float64
-var imageRatioMapMutex sync.RWMutex
-var (
-	audioRatioMap      map[string]float64 = nil
-	audioRatioMapMutex                    = sync.RWMutex{}
-)
-var (
-	audioCompletionRatioMap      map[string]float64 = nil
-	audioCompletionRatioMapMutex                    = sync.RWMutex{}
-)
+var imageRatioMap = types.NewRWMap[string, float64]()
+var audioRatioMap = types.NewRWMap[string, float64]()
+var audioCompletionRatioMap = types.NewRWMap[string, float64]()
 
 func ImageRatio2JSONString() string {
-	imageRatioMapMutex.RLock()
-	defer imageRatioMapMutex.RUnlock()
-	jsonBytes, err := common.Marshal(imageRatioMap)
-	if err != nil {
-		common.SysError("error marshalling cache ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return imageRatioMap.MarshalJSONString()
 }
 
 func UpdateImageRatioByJSONString(jsonStr string) error {
-	imageRatioMapMutex.Lock()
-	defer imageRatioMapMutex.Unlock()
-	imageRatioMap = make(map[string]float64)
-	return common.Unmarshal([]byte(jsonStr), &imageRatioMap)
+	return types.LoadFromJsonString(imageRatioMap, jsonStr)
 }
 
 func GetImageRatio(name string) (float64, bool) {
-	imageRatioMapMutex.RLock()
-	defer imageRatioMapMutex.RUnlock()
-	ratio, ok := imageRatioMap[name]
+	ratio, ok := imageRatioMap.Get(name)
 	if !ok {
 		return 1, false // Default to 1 if not found
 	}
@@ -760,78 +625,31 @@ func GetImageRatio(name string) (float64, bool) {
 }
 
 func AudioRatio2JSONString() string {
-	audioRatioMapMutex.RLock()
-	defer audioRatioMapMutex.RUnlock()
-	jsonBytes, err := common.Marshal(audioRatioMap)
-	if err != nil {
-		common.SysError("error marshalling audio ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return audioRatioMap.MarshalJSONString()
 }
 
 func UpdateAudioRatioByJSONString(jsonStr string) error {
-
-	tmp := make(map[string]float64)
-	if err := common.Unmarshal([]byte(jsonStr), &tmp); err != nil {
-		return err
-	}
-	audioRatioMapMutex.Lock()
-	audioRatioMap = tmp
-	audioRatioMapMutex.Unlock()
-	InvalidateExposedDataCache()
-	return nil
+	return types.LoadFromJsonStringWithCallback(audioRatioMap, jsonStr, InvalidateExposedDataCache)
 }
 
 func AudioCompletionRatio2JSONString() string {
-	audioCompletionRatioMapMutex.RLock()
-	defer audioCompletionRatioMapMutex.RUnlock()
-	jsonBytes, err := common.Marshal(audioCompletionRatioMap)
-	if err != nil {
-		common.SysError("error marshalling audio completion ratio: " + err.Error())
-	}
-	return string(jsonBytes)
+	return audioCompletionRatioMap.MarshalJSONString()
 }
 
 func UpdateAudioCompletionRatioByJSONString(jsonStr string) error {
-	tmp := make(map[string]float64)
-	if err := common.Unmarshal([]byte(jsonStr), &tmp); err != nil {
-		return err
-	}
-	audioCompletionRatioMapMutex.Lock()
-	audioCompletionRatioMap = tmp
-	audioCompletionRatioMapMutex.Unlock()
-	InvalidateExposedDataCache()
-	return nil
+	return types.LoadFromJsonStringWithCallback(audioCompletionRatioMap, jsonStr, InvalidateExposedDataCache)
 }
 
 func GetModelRatioCopy() map[string]float64 {
-	modelRatioMapMutex.RLock()
-	defer modelRatioMapMutex.RUnlock()
-	copyMap := make(map[string]float64, len(modelRatioMap))
-	for k, v := range modelRatioMap {
-		copyMap[k] = v
-	}
-	return copyMap
+	return modelRatioMap.ReadAll()
 }
 
 func GetModelPriceCopy() map[string]float64 {
-	modelPriceMapMutex.RLock()
-	defer modelPriceMapMutex.RUnlock()
-	copyMap := make(map[string]float64, len(modelPriceMap))
-	for k, v := range modelPriceMap {
-		copyMap[k] = v
-	}
-	return copyMap
+	return modelPriceMap.ReadAll()
 }
 
 func GetCompletionRatioCopy() map[string]float64 {
-	CompletionRatioMutex.RLock()
-	defer CompletionRatioMutex.RUnlock()
-	copyMap := make(map[string]float64, len(CompletionRatio))
-	for k, v := range CompletionRatio {
-		copyMap[k] = v
-	}
-	return copyMap
+	return completionRatioMap.ReadAll()
 }
 
 // 转换模型名，减少渠道必须配置各种带参数模型

types/rw_map.go

@@ -80,3 +80,24 @@ func LoadFromJsonString[K comparable, V any](m *RWMap[K, V], jsonStr string) err
 	m.data = make(map[K]V)
 	return common.Unmarshal([]byte(jsonStr), &m.data)
 }
+
+// LoadFromJsonStringWithCallback loads a JSON string into the RWMap and calls the callback on success.
+func LoadFromJsonStringWithCallback[K comparable, V any](m *RWMap[K, V], jsonStr string, onSuccess func()) error {
+	m.mutex.Lock()
+	defer m.mutex.Unlock()
+	m.data = make(map[K]V)
+	err := common.Unmarshal([]byte(jsonStr), &m.data)
+	if err == nil && onSuccess != nil {
+		onSuccess()
+	}
+	return err
+}
+
+// MarshalJSONString returns the JSON string representation of the RWMap.
+func (m *RWMap[K, V]) MarshalJSONString() string {
+	bytes, err := m.MarshalJSON()
+	if err != nil {
+		return "{}"
+	}
+	return string(bytes)
+}

web/src/components/settings/RatioSetting.jsx

@@ -36,6 +36,7 @@ const RatioSetting = () => {
     ModelPrice: '',
     ModelRatio: '',
     CacheRatio: '',
+    CreateCacheRatio: '',
     CompletionRatio: '',
     GroupRatio: '',
     GroupGroupRatio: '',

web/src/components/settings/personal/cards/AccountManagement.jsx

@@ -107,9 +107,11 @@ const AccountManagement = ({
       const res = await API.get('/api/user/oauth/bindings');
       if (res.data.success) {
         setCustomOAuthBindings(res.data.data || []);
+      } else {
+        showError(res.data.message || t('获取绑定信息失败'));
       }
     } catch (error) {
-      // ignore
+      showError(error.response?.data?.message || error.message || t('获取绑定信息失败'));
     }
   };
 
@@ -131,7 +133,7 @@ const AccountManagement = ({
             showError(res.data.message);
           }
         } catch (error) {
-          showError(t('操作失败'));
+          showError(error.response?.data?.message || error.message || t('操作失败'));
         } finally {
           setCustomOAuthLoading((prev) => ({ ...prev, [providerId]: false }));
         }

web/src/components/table/channels/modals/EditChannelModal.jsx

@@ -170,6 +170,7 @@ const EditChannelModal = (props) => {
     allow_service_tier: false,
     disable_store: false, // false = 允许透传（默认开启）
     allow_safety_identifier: false,
+    claude_beta_query: false,
   };
   const [batch, setBatch] = useState(false);
   const [multiToSingle, setMultiToSingle] = useState(false);
@@ -633,6 +634,7 @@ const EditChannelModal = (props) => {
           data.disable_store = parsedSettings.disable_store || false;
           data.allow_safety_identifier =
             parsedSettings.allow_safety_identifier || false;
+          data.claude_beta_query = parsedSettings.claude_beta_query || false;
         } catch (error) {
           console.error('解析其他设置失败:', error);
           data.azure_responses_version = '';
@@ -643,6 +645,7 @@ const EditChannelModal = (props) => {
           data.allow_service_tier = false;
           data.disable_store = false;
           data.allow_safety_identifier = false;
+          data.claude_beta_query = false;
         }
       } else {
         // 兼容历史数据：老渠道没有 settings 时，默认按 json 展示
@@ -652,6 +655,7 @@ const EditChannelModal = (props) => {
         data.allow_service_tier = false;
         data.disable_store = false;
         data.allow_safety_identifier = false;
+        data.claude_beta_query = false;
       }
 
       if (
@@ -1394,6 +1398,9 @@ const EditChannelModal = (props) => {
         settings.allow_safety_identifier =
           localInputs.allow_safety_identifier === true;
       }
+      if (localInputs.type === 14) {
+        settings.claude_beta_query = localInputs.claude_beta_query === true;
+      }
     }
 
     localInputs.settings = JSON.stringify(settings);
@@ -1414,6 +1421,7 @@ const EditChannelModal = (props) => {
     delete localInputs.allow_service_tier;
     delete localInputs.disable_store;
     delete localInputs.allow_safety_identifier;
+    delete localInputs.claude_beta_query;
 
     let res;
     localInputs.auto_ban = localInputs.auto_ban ? 1 : 0;
@@ -3316,6 +3324,24 @@ const EditChannelModal = (props) => {
                       </div>
                     </div>
 
+                    {inputs.type === 14 && (
+                      <Form.Switch
+                        field='claude_beta_query'
+                        label={t('Claude 强制 beta=true')}
+                        checkedText={t('开')}
+                        uncheckedText={t('关')}
+                        onChange={(value) =>
+                          handleChannelOtherSettingsChange(
+                            'claude_beta_query',
+                            value,
+                          )
+                        }
+                        extraText={t(
+                          '开启后，该渠道请求 Claude 时将强制追加 ?beta=true（无需客户端手动传参）',
+                        )}
+                      />
+                    )}
+
                     {inputs.type === 1 && (
                       <Form.Switch
                         field='force_format'

web/src/components/table/channels/modals/ModelTestModal.jsx

@@ -26,8 +26,10 @@ import {
   Tag,
   Typography,
   Select,
+  Switch,
+  Banner,
 } from '@douyinfe/semi-ui';
-import { IconSearch } from '@douyinfe/semi-icons';
+import { IconSearch, IconInfoCircle } from '@douyinfe/semi-icons';
 import { copy, showError, showInfo, showSuccess } from '../../../../helpers';
 import { MODEL_TABLE_PAGE_SIZE } from '../../../../constants';
 
@@ -48,11 +50,25 @@ const ModelTestModal = ({
   setModelTablePage,
   selectedEndpointType,
   setSelectedEndpointType,
+  isStreamTest,
+  setIsStreamTest,
   allSelectingRef,
   isMobile,
   t,
 }) => {
   const hasChannel = Boolean(currentTestChannel);
+  const streamToggleDisabled = [
+    'embeddings',
+    'image-generation',
+    'jina-rerank',
+    'openai-response-compact',
+  ].includes(selectedEndpointType);
+
+  React.useEffect(() => {
+    if (streamToggleDisabled && isStreamTest) {
+      setIsStreamTest(false);
+    }
+  }, [streamToggleDisabled, isStreamTest, setIsStreamTest]);
 
   const filteredModels = hasChannel
     ? currentTestChannel.models
@@ -181,6 +197,7 @@ const ModelTestModal = ({
                 currentTestChannel,
                 record.model,
                 selectedEndpointType,
+                isStreamTest,
               )
             }
             loading={isTesting}
@@ -258,25 +275,46 @@ const ModelTestModal = ({
     >
       {hasChannel && (
         <div className='model-test-scroll'>
-          {/* 端点类型选择器 */}
-          <div className='flex items-center gap-2 w-full mb-2'>
-            <Typography.Text strong>{t('端点类型')}:</Typography.Text>
-            <Select
-              value={selectedEndpointType}
-              onChange={setSelectedEndpointType}
-              optionList={endpointTypeOptions}
-              className='!w-full'
-              placeholder={t('选择端点类型')}
-            />
+          {/* Endpoint toolbar */}
+          <div className='flex flex-col sm:flex-row sm:items-center gap-2 w-full mb-2'>
+            <div className='flex items-center gap-2 flex-1 min-w-0'>
+              <Typography.Text strong className='shrink-0'>
+                {t('端点类型')}:
+              </Typography.Text>
+              <Select
+                value={selectedEndpointType}
+                onChange={setSelectedEndpointType}
+                optionList={endpointTypeOptions}
+                className='!w-full min-w-0'
+                placeholder={t('选择端点类型')}
+              />
+            </div>
+            <div className='flex items-center justify-between sm:justify-end gap-2 shrink-0'>
+              <Typography.Text strong className='shrink-0'>
+                {t('流式')}:
+              </Typography.Text>
+              <Switch
+                checked={isStreamTest}
+                onChange={setIsStreamTest}
+                size='small'
+                disabled={streamToggleDisabled}
+                aria-label={t('流式')}
+              />
+            </div>
           </div>
-          <Typography.Text type='tertiary' size='small' className='block mb-2'>
-            {t(
+
+          <Banner
+            type='info'
+            closeIcon={null}
+            icon={<IconInfoCircle />}
+            className='!rounded-lg mb-2'
+            description={t(
               '说明：本页测试为非流式请求；若渠道仅支持流式返回，可能出现测试失败，请以实际使用为准。',
             )}
-          </Typography.Text>
+          />
 
           {/* 搜索与操作按钮 */}
-          <div className='flex items-center justify-end gap-2 w-full mb-2'>
+          <div className='flex flex-col sm:flex-row sm:items-center gap-2 w-full mb-2'>
             <Input
               placeholder={t('搜索模型...')}
               value={modelSearchKeyword}
@@ -284,16 +322,17 @@ const ModelTestModal = ({
                 setModelSearchKeyword(v);
                 setModelTablePage(1);
               }}
-              className='!w-full'
+              className='!w-full sm:!flex-1'
               prefix={<IconSearch />}
               showClear
             />
 
-            <Button onClick={handleCopySelected}>{t('复制已选')}</Button>
-
-            <Button type='tertiary' onClick={handleSelectSuccess}>
-              {t('选择成功')}
-            </Button>
+            <div className='flex items-center justify-end gap-2'>
+              <Button onClick={handleCopySelected}>{t('复制已选')}</Button>
+              <Button type='tertiary' onClick={handleSelectSuccess}>
+                {t('选择成功')}
+              </Button>
+            </div>
           </div>
 
           <Table

web/src/hooks/channels/useChannelsData.jsx

@@ -87,6 +87,7 @@ export const useChannelsData = () => {
   const [isBatchTesting, setIsBatchTesting] = useState(false);
   const [modelTablePage, setModelTablePage] = useState(1);
   const [selectedEndpointType, setSelectedEndpointType] = useState('');
+  const [isStreamTest, setIsStreamTest] = useState(false);
   const [globalPassThroughEnabled, setGlobalPassThroughEnabled] =
     useState(false);
 
@@ -851,7 +852,12 @@ export const useChannelsData = () => {
   };
 
   // Test channel - 单个模型测试，参考旧版实现
-  const testChannel = async (record, model, endpointType = '') => {
+  const testChannel = async (
+    record,
+    model,
+    endpointType = '',
+    stream = false,
+  ) => {
     const testKey = `${record.id}-${model}`;
 
     // 检查是否应该停止批量测试
@@ -867,6 +873,9 @@ export const useChannelsData = () => {
       if (endpointType) {
         url += `&endpoint_type=${endpointType}`;
       }
+      if (stream) {
+        url += `&stream=true`;
+      }
       const res = await API.get(url);
 
       // 检查是否在请求期间被停止
@@ -995,7 +1004,12 @@ export const useChannelsData = () => {
         );
 
         const batchPromises = batch.map((model) =>
-          testChannel(currentTestChannel, model, selectedEndpointType),
+          testChannel(
+            currentTestChannel,
+            model,
+            selectedEndpointType,
+            isStreamTest,
+          ),
         );
         const batchResults = await Promise.allSettled(batchPromises);
         results.push(...batchResults);
@@ -1080,6 +1094,7 @@ export const useChannelsData = () => {
     setSelectedModelKeys([]);
     setModelTablePage(1);
     setSelectedEndpointType('');
+    setIsStreamTest(false);
     // 可选择性保留测试结果，这里不清空以便用户查看
   };
 
@@ -1170,6 +1185,8 @@ export const useChannelsData = () => {
     setModelTablePage,
     selectedEndpointType,
     setSelectedEndpointType,
+    isStreamTest,
+    setIsStreamTest,
     allSelectingRef,
 
     // Multi-key management states

web/src/i18n/locales/en.json

@@ -1146,6 +1146,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "Tip: {key} in the link will be replaced with the API key, {address} will be replaced with the server address",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "Prompt price: {{symbol}}{{price}} / 1M tokens",
     "提示缓存倍率": "Prompt cache ratio",
+    "缓存创建倍率": "Cache creation ratio",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "Defaults to the 5m cache creation ratio; the 1h cache creation ratio is computed by fixed multiplication (currently 1.6x)",
     "搜索供应商": "Search vendor",
     "搜索关键字": "Search keywords",
     "搜索失败": "Search failed",
@@ -1548,6 +1550,7 @@
     "流": "stream",
     "流式响应完成": "Streaming response completed",
     "流式输出": "Streaming Output",
+    "流式": "Streaming",
     "流量端口": "Traffic Port",
     "浅色": "Light",
     "浅色模式": "Light Mode",

web/src/i18n/locales/fr.json

@@ -1156,6 +1156,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "Astuce : {key} dans le lien sera remplacé par la clé API, {address} sera remplacé par l'adresse du serveur",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "Prix d'invite : {{symbol}}{{price}} / 1M tokens",
     "提示缓存倍率": "Ratio de cache d'invite",
+    "缓存创建倍率": "Ratio de création du cache",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "Par défaut, le ratio de création de cache 5m est utilisé ; le ratio de création de cache 1h est calculé via une multiplication fixe (actuellement 1.6x)",
     "搜索供应商": "Rechercher un fournisseur",
     "搜索关键字": "Rechercher des mots-clés",
     "搜索失败": "Search failed",
@@ -1558,6 +1560,7 @@
     "流": "Flux",
     "流式响应完成": "Flux terminé",
     "流式输出": "Sortie en flux",
+    "流式": "Streaming",
     "流量端口": "Traffic Port",
     "浅色": "Clair",
     "浅色模式": "Mode clair",

web/src/i18n/locales/ja.json

@@ -1141,6 +1141,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "ヒント：リンク内の{key}はAPIキーに、{address}はサーバーURLに置換されます",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "プロンプト料金：{{symbol}}{{price}} / 1M tokens",
     "提示缓存倍率": "プロンプトキャッシュ倍率",
+    "缓存创建倍率": "キャッシュ作成倍率",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "デフォルトは5mのキャッシュ作成倍率です。1hのキャッシュ作成倍率は固定乗数で自動計算されます（現在は1.6倍）",
     "搜索供应商": "プロバイダーで検索",
     "搜索关键字": "検索キーワード",
     "搜索失败": "Search failed",
@@ -1543,6 +1545,7 @@
     "流": "ストリーム",
     "流式响应完成": "ストリーム完了",
     "流式输出": "ストリーム出力",
+    "流式": "ストリーミング",
     "流量端口": "Traffic Port",
     "浅色": "ライト",
     "浅色模式": "ライトモード",

web/src/i18n/locales/ru.json

@@ -1167,6 +1167,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "Промпт: {key} в ссылке будет заменен на API-ключ, {address} будет заменен на адрес сервера",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "Цена промпта: {{symbol}}{{price}} / 1M токенов",
     "提示缓存倍率": "Коэффициент кэша промптов",
+    "缓存创建倍率": "Коэффициент создания кэша",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "По умолчанию используется коэффициент создания кэша 5m; коэффициент создания кэша 1h автоматически вычисляется фиксированным умножением (сейчас 1.6x)",
     "搜索供应商": "Поиск поставщиков",
     "搜索关键字": "Поиск по ключевым словам",
     "搜索失败": "Search failed",
@@ -1569,6 +1571,7 @@
     "流": "Поток",
     "流式响应完成": "Поток завершён",
     "流式输出": "Потоковый вывод",
+    "流式": "Стриминг",
     "流量端口": "Traffic Port",
     "浅色": "Светлая",
     "浅色模式": "Светлый режим",

web/src/i18n/locales/vi.json

@@ -1142,6 +1142,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "Mẹo: {key} trong liên kết sẽ được thay thế bằng khóa API, {address} sẽ được thay thế bằng địa chỉ máy chủ",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "Giá gợi ý: {{symbol}}{{price}} / 1M tokens",
     "提示缓存倍率": "Tỷ lệ bộ nhớ đệm gợi ý",
+    "缓存创建倍率": "Tỷ lệ tạo bộ nhớ đệm",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "Mặc định dùng tỷ lệ tạo bộ nhớ đệm 5m; tỷ lệ tạo bộ nhớ đệm 1h được tự động tính bằng phép nhân cố định (hiện là 1.6x)",
     "搜索供应商": "Tìm kiếm nhà cung cấp",
     "搜索关键字": "Từ khóa tìm kiếm",
     "搜索失败": "Search failed",
@@ -1597,6 +1599,7 @@
     "流": "luồng",
     "流式响应完成": "Luồng hoàn tất",
     "流式输出": "Đầu ra luồng",
+    "流式": "Streaming",
     "流量端口": "Traffic Port",
     "浅色": "Sáng",
     "浅色模式": "Chế độ sáng",

web/src/i18n/locales/zh.json

@@ -1136,6 +1136,8 @@
     "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址": "提示：链接中的{key}将被替换为API密钥，{address}将被替换为服务器地址",
     "提示价格：{{symbol}}{{price}} / 1M tokens": "提示价格：{{symbol}}{{price}} / 1M tokens",
     "提示缓存倍率": "提示缓存倍率",
+    "缓存创建倍率": "缓存创建倍率",
+    "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）": "默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）",
     "搜索供应商": "搜索供应商",
     "搜索关键字": "搜索关键字",
     "搜索失败": "搜索失败",
@@ -1538,6 +1540,7 @@
     "流": "流",
     "流式响应完成": "流式响应完成",
     "流式输出": "流式输出",
+    "流式": "流式",
     "流量端口": "流量端口",
     "浅色": "浅色",
     "浅色模式": "浅色模式",

web/src/pages/Setting/Ratio/ModelRatioSettings.jsx

@@ -43,6 +43,7 @@ export default function ModelRatioSettings(props) {
     ModelPrice: '',
     ModelRatio: '',
     CacheRatio: '',
+    CreateCacheRatio: '',
     CompletionRatio: '',
     ImageRatio: '',
     AudioRatio: '',
@@ -200,6 +201,30 @@ export default function ModelRatioSettings(props) {
             />
           </Col>
         </Row>
+        <Row gutter={16}>
+          <Col xs={24} sm={16}>
+            <Form.TextArea
+              label={t('缓存创建倍率')}
+              extraText={t(
+                '默认为 5m 缓存创建倍率；1h 缓存创建倍率按固定乘法自动计算（当前为 1.6x）',
+              )}
+              placeholder={t('为一个 JSON 文本，键为模型名称，值为倍率')}
+              field={'CreateCacheRatio'}
+              autosize={{ minRows: 6, maxRows: 12 }}
+              trigger='blur'
+              stopValidateWithError
+              rules={[
+                {
+                  validator: (rule, value) => verifyJSON(value),
+                  message: '不是合法的 JSON 字符串',
+                },
+              ]}
+              onChange={(value) =>
+                setInputs({ ...inputs, CreateCacheRatio: value })
+              }
+            />
+          </Col>
+        </Row>
         <Row gutter={16}>
           <Col xs={24} sm={16}>
             <Form.TextArea