chore(deps): bump golang.org/x/oauth2 from 0.21.0 to 0.27.0

Bumps [golang.org/x/oauth2](https://github.com/golang/oauth2) from 0.21.0 to 0.27.0. - [Commits](https://github.com/golang/oauth2/compare/v0.21.0...v0.27.0) --- updated-dependencies: - dependency-name: golang.org/x/oauth2 dependency-version: 0.27.0 dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com>
2025-11-17 13:43:42 +08:00 · 2025-07-18 18:11:50 +00:00
31 changed files with 493 additions and 1217 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -12,5 +12,4 @@ cmd.md
 .env
 /one-api
 temp
-.DS_Store
-/__debug_bin*
+.DS_Store
--- a/2
+++ b/2
@@ -44,4 +44,4 @@ COPY --from=builder2 /build/one-api /

 EXPOSE 3000
 WORKDIR /data
-ENTRYPOINT ["/one-api"]
+ENTRYPOINT ["/one-api"]
--- a/README.md
+++ b/README.md
@@ -385,7 +385,7 @@ graph LR
   + 例子：`NODE_TYPE=slave`
 9. `CHANNEL_UPDATE_FREQUENCY`：设置之后将定期更新渠道余额，单位为分钟，未设置则不进行更新。
   + 例子：`CHANNEL_UPDATE_FREQUENCY=1440`
-10. `CHANNEL_TEST_FREQUENCY`：设置之后将定期检查渠道，单位为分钟，未设置则不进行检查。
+10. `CHANNEL_TEST_FREQUENCY`：设置之后将定期检查渠道，单位为分钟，未设置则不进行检查。 
   +例子：`CHANNEL_TEST_FREQUENCY=1440`
 11. `POLLING_INTERVAL`：批量更新渠道余额以及测试可用性时的请求间隔，单位为秒，默认无间隔。
    + 例子：`POLLING_INTERVAL=5`
--- a/common/config/config.go
+++ b/common/config/config.go
@@ -164,6 +164,3 @@ var UserContentRequestTimeout = env.Int("USER_CONTENT_REQUEST_TIMEOUT", 30)

 var EnforceIncludeUsage = env.Bool("ENFORCE_INCLUDE_USAGE", false)
 var TestPrompt = env.String("TEST_PROMPT", "Output only your specific model name with no additional text.")
-
-// OpenrouterProviderSort is used to determine the order of the providers in the openrouter
-var OpenrouterProviderSort = env.String("OPENROUTER_PROVIDER_SORT", "")
--- a/common/conv/any.go
+++ b/common/conv/any.go
@@ -1,9 +1,6 @@
 package conv

 func AsString(v any) string {
-	if str, ok := v.(string); ok {
-		return str
-	}
-
-	return ""
+	str, _ := v.(string)
+	return str
 }
--- a/controller/channel-test.go
+++ b/controller/channel-test.go
@@ -106,8 +106,6 @@ func testChannel(ctx context.Context, channel *model.Channel, request *relaymode
 	if err != nil {
 		return "", err, nil
 	}
-	c.Set(ctxkey.ConvertedRequest, convertedRequest)
-
 	jsonData, err := json.Marshal(convertedRequest)
 	if err != nil {
 		return "", err, nil
--- a/go.mod
+++ b/go.mod
@@ -1,6 +1,6 @@
 module github.com/songquanpeng/one-api

-go 1.20
+go 1.23.0

 require (
 	cloud.google.com/go/iam v1.1.10
@@ -98,7 +98,7 @@ require (
 	go.opentelemetry.io/otel/trace v1.24.0 // indirect
 	golang.org/x/arch v0.8.0 // indirect
 	golang.org/x/net v0.26.0 // indirect
-	golang.org/x/oauth2 v0.21.0 // indirect
+	golang.org/x/oauth2 v0.27.0 // indirect
 	golang.org/x/sys v0.28.0 // indirect
 	golang.org/x/text v0.21.0 // indirect
 	golang.org/x/time v0.5.0 // indirect
--- a/go.sum
+++ b/go.sum
@@ -73,6 +73,7 @@ github.com/go-logr/logr v1.4.1/go.mod h1:9T104GzyrTigFIr8wt5mBrctHMim0Nb2HLGrmQ4
 github.com/go-logr/stdr v1.2.2 h1:hSWxHoqTgW2S2qGc0LTAI563KZ5YKYRhT3MFKZMbjag=
 github.com/go-logr/stdr v1.2.2/go.mod h1:mMo/vtBO5dYbehREoey6XUKy/eSumjCCveDpRre4VKE=
 github.com/go-playground/assert/v2 v2.2.0 h1:JvknZsQTYeFEAhQwI4qEt9cyV5ONwRHC+lYKSsYSR8s=
+github.com/go-playground/assert/v2 v2.2.0/go.mod h1:VDjEfimB/XKnb+ZQfWdccd7VUvScMdVu0Titje2rxJ4=
 github.com/go-playground/locales v0.14.1 h1:EWaQ/wswjilfKLTECiXz7Rh+3BjFhfDFKv/oXslEjJA=
 github.com/go-playground/locales v0.14.1/go.mod h1:hxrqLVvrK65+Rwrd5Fc6F2O76J/NuW9t0sjnWqG1slY=
 github.com/go-playground/universal-translator v0.18.1 h1:Bcnm0ZwsGyWbCzImXv+pAJnYK9S473LQFuzCbDbfSFY=
@@ -111,8 +112,10 @@ github.com/google/go-cmp v0.4.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/
 github.com/google/go-cmp v0.5.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
 github.com/google/go-cmp v0.5.3/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
 github.com/google/go-cmp v0.6.0 h1:ofyhxvXcZhMsU5ulbFiLKl/XBFqE1GSq7atu8tAmTRI=
+github.com/google/go-cmp v0.6.0/go.mod h1:17dUlkBOakJ0+DkrSSNjCkIjxS6bF9zb3elmeNGIjoY=
 github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
 github.com/google/gofuzz v1.2.0 h1:xRy4A+RhZaiKjJ1bPfwQ8sedCA+YS2YcCHW6ec7JMi0=
+github.com/google/gofuzz v1.2.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
 github.com/google/s2a-go v0.1.7 h1:60BLSyTrOV4/haCDW4zb1guZItoSq8foHCXrAnjBo/o=
 github.com/google/s2a-go v0.1.7/go.mod h1:50CgR4k1jNlWBu4UfS4AcfhVe1r6pdZPygJ3R8F0Qdw=
 github.com/google/uuid v1.1.2/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
@@ -157,6 +160,7 @@ github.com/klauspost/cpuid/v2 v2.2.7 h1:ZWSB3igEs+d0qvnxR/ZBzXVmxkgt8DdzP6m9pfuV
 github.com/klauspost/cpuid/v2 v2.2.7/go.mod h1:Lcz8mBdAVJIBVzewtcLocK12l3Y+JytZYpaMropDUws=
 github.com/knz/go-libedit v1.10.1/go.mod h1:MZTVkCWyz0oBc7JOWP3wNAzd002ZbM/5hgShxwh4x8M=
 github.com/kr/pretty v0.3.0 h1:WgNl7dwNpEZ6jJ9k1snq4pZsg7DOEN8hP9Xw0Tsjwk0=
+github.com/kr/pretty v0.3.0/go.mod h1:640gp4NfQd8pI5XOwp5fnNeVWj67G7CFk/SaSQn7NBk=
 github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/leodido/go-urn v1.4.0 h1:WT9HwE9SGECu3lg4d/dIA+jxlljEa1/ffXKmRjqdmIQ=
@@ -171,8 +175,11 @@ github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd/go.mod h1:6dJ
 github.com/modern-go/reflect2 v1.0.2 h1:xBagoLtFs94CBntxluKeaWgTMpvLxC4ur3nMaC9Gz0M=
 github.com/modern-go/reflect2 v1.0.2/go.mod h1:yWuevngMOJpCy52FWWMvUC8ws7m/LJsjYzDa0/r8luk=
 github.com/nxadm/tail v1.4.8 h1:nPr65rt6Y5JFSKQO7qToXr7pePgD6Gwiw05lkbyAQTE=
+github.com/nxadm/tail v1.4.8/go.mod h1:+ncqLTQzXmGhMZNUePPaPqPvBxHAIsmXswZKocGu+AU=
 github.com/onsi/ginkgo v1.16.5 h1:8xi0RTUf59SOSfEtZMvwTvXYMzG4gV23XVHOZiXNtnE=
+github.com/onsi/ginkgo v1.16.5/go.mod h1:+E8gABHa3K6zRBolWtd+ROzc/U5bkGt0FwiG042wbpU=
 github.com/onsi/gomega v1.18.1 h1:M1GfJqGRrBrrGGsbxzV5dqM2U2ApXefZCQpkukxYRLE=
+github.com/onsi/gomega v1.18.1/go.mod h1:0q+aL8jAiMXy9hbwj2mr5GziHiwhAIQpFmmtT5hitRs=
 github.com/patrickmn/go-cache v2.1.0+incompatible h1:HRMgzkcYKYpi3C8ajMPV8OFXaaRUnok+kx1WdO15EQc=
 github.com/patrickmn/go-cache v2.1.0+incompatible/go.mod h1:3Qf8kWWT7OJRJbdiICTKqZju1ZixQ/KpMGzzAfe6+WQ=
 github.com/pelletier/go-toml/v2 v2.2.2 h1:aYUidT7k73Pcl9nb2gScu7NSrKCSHIDE89b3+6Wq+LM=
@@ -185,6 +192,7 @@ github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZb
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/prometheus/client_model v0.0.0-20190812154241-14fe0d1b01d4/go.mod h1:xMI15A0UPsDsEKsMN9yxemIoYk6Tm2C1GtYGdfGttqA=
 github.com/rogpeppe/go-internal v1.8.0 h1:FCbCCtXNOY3UtUuHUYaghJg4y7Fd14rXifAYUAtL9R8=
+github.com/rogpeppe/go-internal v1.8.0/go.mod h1:WmiCO8CzOY8rg0OYDC4/i/2WRWAB6poM+XZ2dLUbcbE=
 github.com/smarty/assertions v1.15.0 h1:cR//PqUBUiQRakZWqBiFFQ9wb8emQGDb0HeGdqGByCY=
 github.com/smarty/assertions v1.15.0/go.mod h1:yABtdzeQs6l1brC900WlRNwj6ZR55d7B+E8C6HtKdec=
 github.com/smartystreets/goconvey v1.8.1 h1:qGjIddxOk4grTu9JPOU31tVfq3cNdBlNa5sSznIX1xY=
@@ -239,8 +247,8 @@ golang.org/x/net v0.0.0-20201110031124-69a78807bb2b/go.mod h1:sp8m0HH+o8qH0wwXwY
 golang.org/x/net v0.26.0 h1:soB7SVo0PWrY4vPW/+ay0jKDNScG2X9wFeYlXIvJsOQ=
 golang.org/x/net v0.26.0/go.mod h1:5YKkiSynbBIh3p6iOc/vibscux0x38BZDkn8sCUPxHE=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
-golang.org/x/oauth2 v0.21.0 h1:tsimM75w1tF/uws5rbeHzIWxEqElMehnc+iW793zsZs=
-golang.org/x/oauth2 v0.21.0/go.mod h1:XYTD2NtWslqkgxebSiOHnXEap4TF09sJSc7H1sXbhtI=
+golang.org/x/oauth2 v0.27.0 h1:da9Vo7/tDv5RH/7nZDz1eMGS/q1Vv1N/7FCrBhI9I3M=
+golang.org/x/oauth2 v0.27.0/go.mod h1:onh5ek6nERTohokkhCD/y2cV4Do3fxFHFuAejCkRWT8=
 golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20181108010431-42b317875d0f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
@@ -297,8 +305,11 @@ google.golang.org/protobuf v1.34.2 h1:6xV6lTsCfpGD21XK49h7MhtcApnLqkfYgPcdHftf6h
 google.golang.org/protobuf v1.34.2/go.mod h1:qYOHts0dSfpeUzUFpOMr/WGzszTmLH+DiWniOlNbLDw=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
+gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c/go.mod h1:JHkPIbrfpd72SG/EVd6muEfDQjcINNoR0C8j2r3qZ4Q=
 gopkg.in/tomb.v1 v1.0.0-20141024135613-dd632973f1e7 h1:uRGJdciOHaEIrze2W8Q3AKkepLTh2hOroT7a+7czfdQ=
+gopkg.in/tomb.v1 v1.0.0-20141024135613-dd632973f1e7/go.mod h1:dt/ZhP58zS4L8KSrWDmTeBkI65Dw0HsyUHuEVlX15mw=
 gopkg.in/yaml.v2 v2.4.0 h1:D8xgwECY7CYvx+Y2n4sBz93Jn9JRvxdiyyo8CTfuKaY=
+gopkg.in/yaml.v2 v2.4.0/go.mod h1:RDklbk79AGWmwhnvt/jBztapEOGDOx6ZbXqjP6csGnQ=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
 gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
 gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
--- a/relay/adaptor/anthropic/adaptor.go
+++ b/relay/adaptor/anthropic/adaptor.go
@@ -36,8 +36,8 @@ func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *me

 	// https://x.com/alexalbert__/status/1812921642143900036
 	// claude-3-5-sonnet can support 8k context
-	if strings.HasPrefix(meta.ActualModelName, "claude-3-7-sonnet") {
-		req.Header.Set("anthropic-beta", "output-128k-2025-02-19")
+	if strings.HasPrefix(meta.ActualModelName, "claude-3-5-sonnet") {
+		req.Header.Set("anthropic-beta", "max-tokens-3-5-sonnet-2024-07-15")
 	}

 	return nil
@@ -47,7 +47,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	return ConvertRequest(c, *request)
+	return ConvertRequest(*request), nil
 }

 func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
--- a/relay/adaptor/anthropic/constants.go
+++ b/relay/adaptor/anthropic/constants.go
@@ -3,13 +3,11 @@ package anthropic
 var ModelList = []string{
 	"claude-instant-1.2", "claude-2.0", "claude-2.1",
 	"claude-3-haiku-20240307",
-	"claude-3-5-haiku-latest",
 	"claude-3-5-haiku-20241022",
+	"claude-3-5-haiku-latest",
 	"claude-3-sonnet-20240229",
 	"claude-3-opus-20240229",
-	"claude-3-5-sonnet-latest",
 	"claude-3-5-sonnet-20240620",
 	"claude-3-5-sonnet-20241022",
-	"claude-3-7-sonnet-latest",
-	"claude-3-7-sonnet-20250219",
+	"claude-3-5-sonnet-latest",
 }
--- a/relay/adaptor/anthropic/main.go
+++ b/relay/adaptor/anthropic/main.go
@@ -2,21 +2,18 @@ package anthropic

 import (
 	"bufio"
-	"context"
 	"encoding/json"
 	"fmt"
+	"github.com/songquanpeng/one-api/common/render"
 	"io"
-	"math"
 	"net/http"
 	"strings"

 	"github.com/gin-gonic/gin"
-	"github.com/pkg/errors"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/image"
 	"github.com/songquanpeng/one-api/common/logger"
-	"github.com/songquanpeng/one-api/common/render"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	"github.com/songquanpeng/one-api/relay/model"
 )
@@ -39,16 +36,7 @@ func stopReasonClaude2OpenAI(reason *string) string {
 	}
 }

-// isModelSupportThinking is used to check if the model supports extended thinking
-func isModelSupportThinking(model string) bool {
-	if strings.Contains(model, "claude-3-7-sonnet") {
-		return true
-	}
-
-	return false
-}
-
-func ConvertRequest(c *gin.Context, textRequest model.GeneralOpenAIRequest) (*Request, error) {
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
 	claudeTools := make([]Tool, 0, len(textRequest.Tools))

 	for _, tool := range textRequest.Tools {
@@ -73,27 +61,7 @@ func ConvertRequest(c *gin.Context, textRequest model.GeneralOpenAIRequest) (*Re
 		TopK:        textRequest.TopK,
 		Stream:      textRequest.Stream,
 		Tools:       claudeTools,
-		Thinking:    textRequest.Thinking,
 	}
-
-	if isModelSupportThinking(textRequest.Model) &&
-		c.Request.URL.Query().Has("thinking") && claudeRequest.Thinking == nil {
-		claudeRequest.Thinking = &model.Thinking{
-			Type:         "enabled",
-			BudgetTokens: int(math.Min(1024, float64(claudeRequest.MaxTokens/2))),
-		}
-	}
-
-	if isModelSupportThinking(textRequest.Model) &&
-		claudeRequest.Thinking != nil {
-		if claudeRequest.MaxTokens <= 1024 {
-			return nil, errors.New("max_tokens must be greater than 1024 when using extended thinking")
-		}
-
-		// top_p must be nil when using extended thinking
-		claudeRequest.TopP = nil
-	}
-
 	if len(claudeTools) > 0 {
 		claudeToolChoice := struct {
 			Type string `json:"type"`
@@ -174,14 +142,13 @@ func ConvertRequest(c *gin.Context, textRequest model.GeneralOpenAIRequest) (*Re
 		claudeMessage.Content = contents
 		claudeRequest.Messages = append(claudeRequest.Messages, claudeMessage)
 	}
-	return &claudeRequest, nil
+	return &claudeRequest
 }

 // https://docs.anthropic.com/claude/reference/messages-streaming
 func StreamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCompletionsStreamResponse, *Response) {
 	var response *Response
 	var responseText string
-	var reasoningText string
 	var stopReason string
 	tools := make([]model.Tool, 0)

@@ -191,10 +158,6 @@ func StreamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCo
 	case "content_block_start":
 		if claudeResponse.ContentBlock != nil {
 			responseText = claudeResponse.ContentBlock.Text
-			if claudeResponse.ContentBlock.Thinking != nil {
-				reasoningText = *claudeResponse.ContentBlock.Thinking
-			}
-
 			if claudeResponse.ContentBlock.Type == "tool_use" {
 				tools = append(tools, model.Tool{
 					Id:   claudeResponse.ContentBlock.Id,
@@ -209,10 +172,6 @@ func StreamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCo
 	case "content_block_delta":
 		if claudeResponse.Delta != nil {
 			responseText = claudeResponse.Delta.Text
-			if claudeResponse.Delta.Thinking != nil {
-				reasoningText = *claudeResponse.Delta.Thinking
-			}
-
 			if claudeResponse.Delta.Type == "input_json_delta" {
 				tools = append(tools, model.Tool{
 					Function: model.Function{
@@ -230,20 +189,9 @@ func StreamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCo
 		if claudeResponse.Delta != nil && claudeResponse.Delta.StopReason != nil {
 			stopReason = *claudeResponse.Delta.StopReason
 		}
-	case "thinking_delta":
-		if claudeResponse.Delta != nil && claudeResponse.Delta.Thinking != nil {
-			reasoningText = *claudeResponse.Delta.Thinking
-		}
-	case "ping",
-		"message_stop",
-		"content_block_stop":
-	default:
-		logger.SysErrorf("unknown stream response type %q", claudeResponse.Type)
 	}
-
 	var choice openai.ChatCompletionsStreamResponseChoice
 	choice.Delta.Content = responseText
-	choice.Delta.Reasoning = &reasoningText
 	if len(tools) > 0 {
 		choice.Delta.Content = nil // compatible with other OpenAI derivative applications, like LobeOpenAICompatibleFactory ...
 		choice.Delta.ToolCalls = tools
@@ -261,23 +209,11 @@ func StreamResponseClaude2OpenAI(claudeResponse *StreamResponse) (*openai.ChatCo

 func ResponseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
 	var responseText string
-	var reasoningText string
-
+	if len(claudeResponse.Content) > 0 {
+		responseText = claudeResponse.Content[0].Text
+	}
 	tools := make([]model.Tool, 0)
 	for _, v := range claudeResponse.Content {
-		switch v.Type {
-		case "thinking":
-			if v.Thinking != nil {
-				reasoningText += *v.Thinking
-			} else {
-				logger.Errorf(context.Background(), "thinking is nil in response")
-			}
-		case "text":
-			responseText += v.Text
-		default:
-			logger.Warnf(context.Background(), "unknown response type %q", v.Type)
-		}
-
 		if v.Type == "tool_use" {
 			args, _ := json.Marshal(v.Input)
 			tools = append(tools, model.Tool{
@@ -290,13 +226,11 @@ func ResponseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
 			})
 		}
 	}
-
 	choice := openai.TextResponseChoice{
 		Index: 0,
 		Message: model.Message{
 			Role:      "assistant",
 			Content:   responseText,
-			Reasoning: &reasoningText,
 			Name:      nil,
 			ToolCalls: tools,
 		},
@@ -343,8 +277,6 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusC
 		data = strings.TrimPrefix(data, "data:")
 		data = strings.TrimSpace(data)

-		logger.Debugf(c.Request.Context(), "stream <- %q\n", data)
-
 		var claudeResponse StreamResponse
 		err := json.Unmarshal([]byte(data), &claudeResponse)
 		if err != nil {
@@ -412,9 +344,6 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 	if err != nil {
 		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
-
-	logger.Debugf(c.Request.Context(), "response <- %s\n", string(responseBody))
-
 	var claudeResponse Response
 	err = json.Unmarshal(responseBody, &claudeResponse)
 	if err != nil {
--- a/relay/adaptor/anthropic/model.go
+++ b/relay/adaptor/anthropic/model.go
@@ -1,7 +1,5 @@
 package anthropic

-import "github.com/songquanpeng/one-api/relay/model"
-
 // https://docs.anthropic.com/claude/reference/messages_post

 type Metadata struct {
@@ -24,9 +22,6 @@ type Content struct {
 	Input     any    `json:"input,omitempty"`
 	Content   string `json:"content,omitempty"`
 	ToolUseId string `json:"tool_use_id,omitempty"`
-	// https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#implementing-extended-thinking
-	Thinking  *string `json:"thinking,omitempty"`
-	Signature *string `json:"signature,omitempty"`
 }

 type Message struct {
@@ -59,7 +54,6 @@ type Request struct {
 	Tools         []Tool    `json:"tools,omitempty"`
 	ToolChoice    any       `json:"tool_choice,omitempty"`
 	//Metadata    `json:"metadata,omitempty"`
-	Thinking *model.Thinking `json:"thinking,omitempty"`
 }

 type Usage struct {
@@ -90,8 +84,6 @@ type Delta struct {
 	PartialJson  string  `json:"partial_json,omitempty"`
 	StopReason   *string `json:"stop_reason"`
 	StopSequence *string `json:"stop_sequence"`
-	Thinking     *string `json:"thinking,omitempty"`
-	Signature    *string `json:"signature,omitempty"`
 }

 type StreamResponse struct {
--- a/relay/adaptor/aws/claude/adapter.go
+++ b/relay/adaptor/aws/claude/adapter.go
@@ -21,11 +21,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 		return nil, errors.New("request is nil")
 	}

-	claudeReq, err := anthropic.ConvertRequest(c, *request)
-	if err != nil {
-		return nil, errors.Wrap(err, "convert request")
-	}
-
+	claudeReq := anthropic.ConvertRequest(*request)
 	c.Set(ctxkey.RequestModel, request.Model)
 	c.Set(ctxkey.ConvertedRequest, claudeReq)
 	return claudeReq, nil
--- a/relay/adaptor/aws/claude/main.go
+++ b/relay/adaptor/aws/claude/main.go
@@ -36,8 +36,6 @@ var AwsModelIDMap = map[string]string{
 	"claude-3-5-sonnet-20241022": "anthropic.claude-3-5-sonnet-20241022-v2:0",
 	"claude-3-5-sonnet-latest":   "anthropic.claude-3-5-sonnet-20241022-v2:0",
 	"claude-3-5-haiku-20241022":  "anthropic.claude-3-5-haiku-20241022-v1:0",
-	"claude-3-7-sonnet-latest":   "anthropic.claude-3-7-sonnet-20250219-v1:0",
-	"claude-3-7-sonnet-20250219": "anthropic.claude-3-7-sonnet-20250219-v1:0",
 }

 func awsModelID(requestModel string) (string, error) {
@@ -49,14 +47,13 @@ func awsModelID(requestModel string) (string, error) {
 }

 func Handler(c *gin.Context, awsCli *bedrockruntime.Client, modelName string) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
-	awsModelID, err := awsModelID(c.GetString(ctxkey.RequestModel))
+	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return utils.WrapErr(errors.Wrap(err, "awsModelID")), nil
 	}

-	awsModelID = utils.ConvertModelID2CrossRegionProfile(awsModelID, awsCli.Options().Region)
 	awsReq := &bedrockruntime.InvokeModelInput{
-		ModelId:     aws.String(awsModelID),
+		ModelId:     aws.String(awsModelId),
 		Accept:      aws.String("application/json"),
 		ContentType: aws.String("application/json"),
 	}
@@ -104,14 +101,13 @@ func Handler(c *gin.Context, awsCli *bedrockruntime.Client, modelName string) (*

 func StreamHandler(c *gin.Context, awsCli *bedrockruntime.Client) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
 	createdTime := helper.GetTimestamp()
-	awsModelID, err := awsModelID(c.GetString(ctxkey.RequestModel))
+	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return utils.WrapErr(errors.Wrap(err, "awsModelID")), nil
 	}

-	awsModelID = utils.ConvertModelID2CrossRegionProfile(awsModelID, awsCli.Options().Region)
 	awsReq := &bedrockruntime.InvokeModelWithResponseStreamInput{
-		ModelId:     aws.String(awsModelID),
+		ModelId:     aws.String(awsModelId),
 		Accept:      aws.String("application/json"),
 		ContentType: aws.String("application/json"),
 	}
--- a/relay/adaptor/aws/claude/model.go
+++ b/relay/adaptor/aws/claude/model.go
@@ -1,9 +1,6 @@
 package aws

-import (
-	"github.com/songquanpeng/one-api/relay/adaptor/anthropic"
-	"github.com/songquanpeng/one-api/relay/model"
-)
+import "github.com/songquanpeng/one-api/relay/adaptor/anthropic"

 // Request is the request to AWS Claude
 //
@@ -20,5 +17,4 @@ type Request struct {
 	StopSequences    []string            `json:"stop_sequences,omitempty"`
 	Tools            []anthropic.Tool    `json:"tools,omitempty"`
 	ToolChoice       any                 `json:"tool_choice,omitempty"`
-	Thinking         *model.Thinking     `json:"thinking,omitempty"`
 }
--- a/relay/adaptor/aws/llama3/main.go
+++ b/relay/adaptor/aws/llama3/main.go
@@ -70,14 +70,13 @@ func ConvertRequest(textRequest relaymodel.GeneralOpenAIRequest) *Request {
 }

 func Handler(c *gin.Context, awsCli *bedrockruntime.Client, modelName string) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
-	awsModelID, err := awsModelID(c.GetString(ctxkey.RequestModel))
+	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return utils.WrapErr(errors.Wrap(err, "awsModelID")), nil
 	}

-	awsModelID = utils.ConvertModelID2CrossRegionProfile(awsModelID, awsCli.Options().Region)
 	awsReq := &bedrockruntime.InvokeModelInput{
-		ModelId:     aws.String(awsModelID),
+		ModelId:     aws.String(awsModelId),
 		Accept:      aws.String("application/json"),
 		ContentType: aws.String("application/json"),
 	}
@@ -141,14 +140,13 @@ func ResponseLlama2OpenAI(llamaResponse *Response) *openai.TextResponse {

 func StreamHandler(c *gin.Context, awsCli *bedrockruntime.Client) (*relaymodel.ErrorWithStatusCode, *relaymodel.Usage) {
 	createdTime := helper.GetTimestamp()
-	awsModelID, err := awsModelID(c.GetString(ctxkey.RequestModel))
+	awsModelId, err := awsModelID(c.GetString(ctxkey.RequestModel))
 	if err != nil {
 		return utils.WrapErr(errors.Wrap(err, "awsModelID")), nil
 	}

-	awsModelID = utils.ConvertModelID2CrossRegionProfile(awsModelID, awsCli.Options().Region)
 	awsReq := &bedrockruntime.InvokeModelWithResponseStreamInput{
-		ModelId:     aws.String(awsModelID),
+		ModelId:     aws.String(awsModelId),
 		Accept:      aws.String("application/json"),
 		ContentType: aws.String("application/json"),
 	}
--- a/relay/adaptor/aws/utils/consts.go
+++ b/relay/adaptor/aws/utils/consts.go
@@ -1,75 +0,0 @@
-package utils
-
-import (
-	"context"
-	"slices"
-	"strings"
-
-	"github.com/songquanpeng/one-api/common/logger"
-)
-
-// CrossRegionInferences is a list of model IDs that support cross-region inference.
-//
-// https://docs.aws.amazon.com/bedrock/latest/userguide/inference-profiles-support.html
-//
-// document.querySelectorAll('pre.programlisting code').forEach((e) => {console.log(e.innerHTML)})
-var CrossRegionInferences = []string{
-	"us.amazon.nova-lite-v1:0",
-	"us.amazon.nova-micro-v1:0",
-	"us.amazon.nova-pro-v1:0",
-	"us.anthropic.claude-3-5-haiku-20241022-v1:0",
-	"us.anthropic.claude-3-5-sonnet-20240620-v1:0",
-	"us.anthropic.claude-3-5-sonnet-20241022-v2:0",
-	"us.anthropic.claude-3-7-sonnet-20250219-v1:0",
-	"us.anthropic.claude-3-haiku-20240307-v1:0",
-	"us.anthropic.claude-3-opus-20240229-v1:0",
-	"us.anthropic.claude-3-sonnet-20240229-v1:0",
-	"us.meta.llama3-1-405b-instruct-v1:0",
-	"us.meta.llama3-1-70b-instruct-v1:0",
-	"us.meta.llama3-1-8b-instruct-v1:0",
-	"us.meta.llama3-2-11b-instruct-v1:0",
-	"us.meta.llama3-2-1b-instruct-v1:0",
-	"us.meta.llama3-2-3b-instruct-v1:0",
-	"us.meta.llama3-2-90b-instruct-v1:0",
-	"us.meta.llama3-3-70b-instruct-v1:0",
-	"us-gov.anthropic.claude-3-5-sonnet-20240620-v1:0",
-	"us-gov.anthropic.claude-3-haiku-20240307-v1:0",
-	"eu.amazon.nova-lite-v1:0",
-	"eu.amazon.nova-micro-v1:0",
-	"eu.amazon.nova-pro-v1:0",
-	"eu.anthropic.claude-3-5-sonnet-20240620-v1:0",
-	"eu.anthropic.claude-3-haiku-20240307-v1:0",
-	"eu.anthropic.claude-3-sonnet-20240229-v1:0",
-	"eu.meta.llama3-2-1b-instruct-v1:0",
-	"eu.meta.llama3-2-3b-instruct-v1:0",
-	"apac.amazon.nova-lite-v1:0",
-	"apac.amazon.nova-micro-v1:0",
-	"apac.amazon.nova-pro-v1:0",
-	"apac.anthropic.claude-3-5-sonnet-20240620-v1:0",
-	"apac.anthropic.claude-3-5-sonnet-20241022-v2:0",
-	"apac.anthropic.claude-3-haiku-20240307-v1:0",
-	"apac.anthropic.claude-3-sonnet-20240229-v1:0",
-}
-
-// ConvertModelID2CrossRegionProfile converts the model ID to a cross-region profile ID.
-func ConvertModelID2CrossRegionProfile(model, region string) string {
-	var regionPrefix string
-	switch prefix := strings.Split(region, "-")[0]; prefix {
-	case "us", "eu":
-		regionPrefix = prefix
-	case "ap":
-		regionPrefix = "apac"
-	default:
-		// not supported, return original model
-		return model
-	}
-
-	newModelID := regionPrefix + "." + model
-	if slices.Contains(CrossRegionInferences, newModelID) {
-		logger.Debugf(context.TODO(), "convert model %s to cross-region profile %s", model, newModelID)
-		return newModelID
-	}
-
-	// not found, return original model
-	return model
-}
--- a/relay/adaptor/cloudflare/main.go
+++ b/relay/adaptor/cloudflare/main.go
@@ -19,8 +19,9 @@ import (
 )

 func ConvertCompletionsRequest(textRequest model.GeneralOpenAIRequest) *Request {
+	p, _ := textRequest.Prompt.(string)
 	return &Request{
-		Prompt:      textRequest.Prompt,
+		Prompt:      p,
 		MaxTokens:   textRequest.MaxTokens,
 		Stream:      textRequest.Stream,
 		Temperature: textRequest.Temperature,
--- a/relay/adaptor/openai/adaptor.go
+++ b/relay/adaptor/openai/adaptor.go
@@ -1,18 +1,14 @@
 package openai

 import (
+	"errors"
 	"fmt"
 	"io"
-	"math"
 	"net/http"
 	"strings"

 	"github.com/gin-gonic/gin"
-	"github.com/pkg/errors"

-	"github.com/songquanpeng/one-api/common/config"
-	"github.com/songquanpeng/one-api/common/ctxkey"
-	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay/adaptor"
 	"github.com/songquanpeng/one-api/relay/adaptor/alibailian"
 	"github.com/songquanpeng/one-api/relay/adaptor/baiduv2"
@@ -20,8 +16,6 @@ import (
 	"github.com/songquanpeng/one-api/relay/adaptor/geminiv2"
 	"github.com/songquanpeng/one-api/relay/adaptor/minimax"
 	"github.com/songquanpeng/one-api/relay/adaptor/novita"
-	"github.com/songquanpeng/one-api/relay/adaptor/openrouter"
-	"github.com/songquanpeng/one-api/relay/billing/ratio"
 	"github.com/songquanpeng/one-api/relay/channeltype"
 	"github.com/songquanpeng/one-api/relay/meta"
 	"github.com/songquanpeng/one-api/relay/model"
@@ -91,72 +85,13 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-
-	meta := meta.GetByContext(c)
-	switch meta.ChannelType {
-	case channeltype.OpenRouter:
-		includeReasoning := true
-		request.IncludeReasoning = &includeReasoning
-		if request.Provider == nil || request.Provider.Sort == "" &&
-			config.OpenrouterProviderSort != "" {
-			if request.Provider == nil {
-				request.Provider = &openrouter.RequestProvider{}
-			}
-
-			request.Provider.Sort = config.OpenrouterProviderSort
-		}
-	default:
-	}
-
-	if request.Stream && !config.EnforceIncludeUsage {
-		logger.Warn(c.Request.Context(),
-			"please set ENFORCE_INCLUDE_USAGE=true to ensure accurate billing in stream mode")
-	}
-
-	if config.EnforceIncludeUsage && request.Stream {
+	if request.Stream {
 		// always return usage in stream mode
 		if request.StreamOptions == nil {
 			request.StreamOptions = &model.StreamOptions{}
 		}
 		request.StreamOptions.IncludeUsage = true
 	}
-
-	// o1/o1-mini/o1-preview do not support system prompt/max_tokens/temperature
-	if strings.HasPrefix(meta.ActualModelName, "o1") ||
-		strings.HasPrefix(meta.ActualModelName, "o3") {
-		temperature := float64(1)
-		request.Temperature = &temperature // Only the default (1) value is supported
-
-		request.MaxTokens = 0
-		request.Messages = func(raw []model.Message) (filtered []model.Message) {
-			for i := range raw {
-				if raw[i].Role != "system" {
-					filtered = append(filtered, raw[i])
-				}
-			}
-
-			return
-		}(request.Messages)
-	}
-
-	// web search do not support system prompt/max_tokens/temperature
-	if strings.HasPrefix(meta.ActualModelName, "gpt-4o-search") ||
-		strings.HasPrefix(meta.ActualModelName, "gpt-4o-mini-search") {
-		request.Temperature = nil
-		request.TopP = nil
-		request.PresencePenalty = nil
-		request.N = nil
-		request.FrequencyPenalty = nil
-	}
-
-	if request.Stream && !config.EnforceIncludeUsage &&
-		(strings.HasPrefix(request.Model, "gpt-4o-audio") ||
-			strings.HasPrefix(request.Model, "gpt-4o-mini-audio")) {
-		// TODO: Since it is not clear how to implement billing in stream mode,
-		// it is temporarily not supported
-		return nil, errors.New("set ENFORCE_INCLUDE_USAGE=true to enable stream mode for gpt-4o-audio")
-	}
-
 	return request, nil
 }

@@ -167,16 +102,11 @@ func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error)
 	return request, nil
 }

-func (a *Adaptor) DoRequest(c *gin.Context,
-	meta *meta.Meta,
-	requestBody io.Reader) (*http.Response, error) {
+func (a *Adaptor) DoRequest(c *gin.Context, meta *meta.Meta, requestBody io.Reader) (*http.Response, error) {
 	return adaptor.DoRequestHelper(a, c, meta, requestBody)
 }

-func (a *Adaptor) DoResponse(c *gin.Context,
-	resp *http.Response,
-	meta *meta.Meta) (usage *model.Usage,
-	err *model.ErrorWithStatusCode) {
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *meta.Meta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
 	if meta.IsStream {
 		var responseText string
 		err, responseText, usage = StreamHandler(c, resp, meta.Mode)
@@ -195,55 +125,6 @@ func (a *Adaptor) DoResponse(c *gin.Context,
 			err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
 		}
 	}
-
-	// -------------------------------------
-	// calculate web-search tool cost
-	// -------------------------------------
-	if usage != nil {
-		searchContextSize := "medium"
-		var req *model.GeneralOpenAIRequest
-		if vi, ok := c.Get(ctxkey.ConvertedRequest); ok {
-			if req, ok = vi.(*model.GeneralOpenAIRequest); ok {
-				if req != nil &&
-					req.WebSearchOptions != nil &&
-					req.WebSearchOptions.SearchContextSize != nil {
-					searchContextSize = *req.WebSearchOptions.SearchContextSize
-				}
-
-				switch {
-				case strings.HasPrefix(meta.ActualModelName, "gpt-4o-search"):
-					switch searchContextSize {
-					case "low":
-						usage.ToolsCost += int64(math.Ceil(30 / 1000 * ratio.QuotaPerUsd))
-					case "medium":
-						usage.ToolsCost += int64(math.Ceil(35 / 1000 * ratio.QuotaPerUsd))
-					case "high":
-						usage.ToolsCost += int64(math.Ceil(40 / 1000 * ratio.QuotaPerUsd))
-					default:
-						return nil, ErrorWrapper(
-							errors.Errorf("invalid search context size %q", searchContextSize),
-							"invalid search context size: "+searchContextSize,
-							http.StatusBadRequest)
-					}
-				case strings.HasPrefix(meta.ActualModelName, "gpt-4o-mini-search"):
-					switch searchContextSize {
-					case "low":
-						usage.ToolsCost += int64(math.Ceil(25 / 1000 * ratio.QuotaPerUsd))
-					case "medium":
-						usage.ToolsCost += int64(math.Ceil(27.5 / 1000 * ratio.QuotaPerUsd))
-					case "high":
-						usage.ToolsCost += int64(math.Ceil(30 / 1000 * ratio.QuotaPerUsd))
-					default:
-						return nil, ErrorWrapper(
-							errors.Errorf("invalid search context size %q", searchContextSize),
-							"invalid search context size: "+searchContextSize,
-							http.StatusBadRequest)
-					}
-				}
-			}
-		}
-	}
-
 	return
 }

--- a/relay/adaptor/openai/constants.go
+++ b/relay/adaptor/openai/constants.go
@@ -24,8 +24,4 @@ var ModelList = []string{
 	"o1", "o1-2024-12-17",
 	"o1-preview", "o1-preview-2024-09-12",
 	"o1-mini", "o1-mini-2024-09-12",
-	"o3-mini", "o3-mini-2025-01-31",
-	"gpt-4.5-preview", "gpt-4.5-preview-2025-02-27",
-	// https://platform.openai.com/docs/guides/tools-web-search?api-mode=chat
-	"gpt-4o-search-preview", "gpt-4o-mini-search-preview",
 }
--- a/relay/adaptor/openai/main.go
+++ b/relay/adaptor/openai/main.go
@@ -8,11 +8,12 @@ import (
 	"net/http"
 	"strings"

+	"github.com/songquanpeng/one-api/common/render"
+
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/conv"
 	"github.com/songquanpeng/one-api/common/logger"
-	"github.com/songquanpeng/one-api/common/render"
 	"github.com/songquanpeng/one-api/relay/model"
 	"github.com/songquanpeng/one-api/relay/relaymode"
 )
@@ -25,7 +26,6 @@ const (

 func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.ErrorWithStatusCode, string, *model.Usage) {
 	responseText := ""
-	reasoningText := ""
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(bufio.ScanLines)
 	var usage *model.Usage
@@ -61,13 +61,6 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.E
 			}
 			render.StringData(c, data)
 			for _, choice := range streamResponse.Choices {
-				if choice.Delta.Reasoning != nil {
-					reasoningText += *choice.Delta.Reasoning
-				}
-				if choice.Delta.ReasoningContent != nil {
-					reasoningText += *choice.Delta.ReasoningContent
-				}
-
 				responseText += conv.AsString(choice.Delta.Content)
 			}
 			if streamResponse.Usage != nil {
@@ -100,7 +93,7 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.E
 		return ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), "", nil
 	}

-	return nil, reasoningText + responseText, usage
+	return nil, responseText, usage
 }

 func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
@@ -143,17 +136,10 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 		return ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}

-	if textResponse.Usage.TotalTokens == 0 ||
-		(textResponse.Usage.PromptTokens == 0 && textResponse.Usage.CompletionTokens == 0) {
+	if textResponse.Usage.TotalTokens == 0 || (textResponse.Usage.PromptTokens == 0 && textResponse.Usage.CompletionTokens == 0) {
 		completionTokens := 0
 		for _, choice := range textResponse.Choices {
 			completionTokens += CountTokenText(choice.Message.StringContent(), modelName)
-			if choice.Message.Reasoning != nil {
-				completionTokens += CountToken(*choice.Message.Reasoning)
-			}
-			if choice.ReasoningContent != nil {
-				completionTokens += CountToken(*choice.ReasoningContent)
-			}
 		}
 		textResponse.Usage = model.Usage{
 			PromptTokens:     promptTokens,
@@ -161,6 +147,5 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName st
 			TotalTokens:      promptTokens + completionTokens,
 		}
 	}
-
 	return nil, &textResponse.Usage
 }
--- a/relay/adaptor/openrouter/model.go
+++ b/relay/adaptor/openrouter/model.go
@@ -1,22 +0,0 @@
-package openrouter
-
-// RequestProvider customize how your requests are routed using the provider object
-// in the request body for Chat Completions and Completions.
-//
-// https://openrouter.ai/docs/features/provider-routing
-type RequestProvider struct {
-	// Order is list of provider names to try in order (e.g. ["Anthropic", "OpenAI"]). Default: empty
-	Order []string `json:"order,omitempty"`
-	// AllowFallbacks is whether to allow backup providers when the primary is unavailable. Default: true
-	AllowFallbacks bool `json:"allow_fallbacks,omitempty"`
-	// RequireParameters is only use providers that support all parameters in your request. Default: false
-	RequireParameters bool `json:"require_parameters,omitempty"`
-	// DataCollection is control whether to use providers that may store data ("allow" or "deny"). Default: "allow"
-	DataCollection string `json:"data_collection,omitempty" binding:"omitempty,oneof=allow deny"`
-	// Ignore is list of provider names to skip for this request. Default: empty
-	Ignore []string `json:"ignore,omitempty"`
-	// Quantizations is list of quantization levels to filter by (e.g. ["int4", "int8"]). Default: empty
-	Quantizations []string `json:"quantizations,omitempty"`
-	// Sort is sort providers by price or throughput (e.g. "price" or "throughput"). Default: empty
-	Sort string `json:"sort,omitempty" binding:"omitempty,oneof=price throughput latency"`
-}
--- a/relay/adaptor/palm/palm.go
+++ b/relay/adaptor/palm/palm.go
@@ -25,17 +25,11 @@ func ConvertRequest(textRequest model.GeneralOpenAIRequest) *ChatRequest {
 		Prompt: Prompt{
 			Messages: make([]ChatMessage, 0, len(textRequest.Messages)),
 		},
-		Temperature: textRequest.Temperature,
-		TopP:        textRequest.TopP,
-		TopK:        textRequest.MaxTokens,
+		Temperature:    textRequest.Temperature,
+		CandidateCount: textRequest.N,
+		TopP:           textRequest.TopP,
+		TopK:           textRequest.MaxTokens,
 	}
-
-	if textRequest.N != nil {
-		palmRequest.CandidateCount = *textRequest.N
-	} else {
-		palmRequest.CandidateCount = 1
-	}
-
 	for _, message := range textRequest.Messages {
 		palmMessage := ChatMessage{
 			Content: message.StringContent(),
--- a/relay/adaptor/vertexai/claude/adapter.go
+++ b/relay/adaptor/vertexai/claude/adapter.go
@@ -19,7 +19,6 @@ var ModelList = []string{
 	"claude-3-5-sonnet@20240620",
 	"claude-3-5-sonnet-v2@20241022",
 	"claude-3-5-haiku@20241022",
-	"claude-3-7-sonnet@20250219",
 }

 const anthropicVersion = "vertex-2023-10-16"
@@ -32,11 +31,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 		return nil, errors.New("request is nil")
 	}

-	claudeReq, err := anthropic.ConvertRequest(c, *request)
-	if err != nil {
-		return nil, errors.Wrap(err, "convert request")
-	}
-
+	claudeReq := anthropic.ConvertRequest(*request)
 	req := Request{
 		AnthropicVersion: anthropicVersion,
 		// Model:            claudeReq.Model,
--- a/relay/adaptor/xunfei/main.go
+++ b/relay/adaptor/xunfei/main.go
@@ -41,15 +41,10 @@ func requestOpenAI2Xunfei(request model.GeneralOpenAIRequest, xunfeiAppId string
 	xunfeiRequest.Header.AppId = xunfeiAppId
 	xunfeiRequest.Parameter.Chat.Domain = domain
 	xunfeiRequest.Parameter.Chat.Temperature = request.Temperature
+	xunfeiRequest.Parameter.Chat.TopK = request.N
 	xunfeiRequest.Parameter.Chat.MaxTokens = request.MaxTokens
 	xunfeiRequest.Payload.Message.Text = messages

-	if request.N != nil {
-		xunfeiRequest.Parameter.Chat.TopK = *request.N
-	} else {
-		xunfeiRequest.Parameter.Chat.TopK = 1
-	}
-
 	if strings.HasPrefix(domain, "generalv3") || domain == "4.0Ultra" {
 		functions := make([]model.Function, len(request.Tools))
 		for i, tool := range request.Tools {
--- a/relay/billing/ratio/model.go
+++ b/relay/billing/ratio/model.go
--- a/relay/controller/helper.go
+++ b/relay/controller/helper.go
@@ -8,16 +8,18 @@ import (
 	"net/http"
 	"strings"

+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/relay/constant/role"
+
 	"github.com/gin-gonic/gin"
+
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/config"
-	"github.com/songquanpeng/one-api/common/helper"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/model"
 	"github.com/songquanpeng/one-api/relay/adaptor/openai"
 	billingratio "github.com/songquanpeng/one-api/relay/billing/ratio"
 	"github.com/songquanpeng/one-api/relay/channeltype"
-	"github.com/songquanpeng/one-api/relay/constant/role"
 	"github.com/songquanpeng/one-api/relay/controller/validator"
 	"github.com/songquanpeng/one-api/relay/meta"
 	relaymodel "github.com/songquanpeng/one-api/relay/model"
@@ -92,30 +94,19 @@ func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIR
 	return preConsumedQuota, nil
 }

-func postConsumeQuota(ctx context.Context,
-	usage *relaymodel.Usage,
-	meta *meta.Meta,
-	textRequest *relaymodel.GeneralOpenAIRequest,
-	ratio float64,
-	preConsumedQuota int64,
-	modelRatio float64,
-	groupRatio float64,
-	systemPromptReset bool) (quota int64) {
+func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *meta.Meta, textRequest *relaymodel.GeneralOpenAIRequest, ratio float64, preConsumedQuota int64, modelRatio float64, groupRatio float64, systemPromptReset bool) {
 	if usage == nil {
 		logger.Error(ctx, "usage is nil, which is unexpected")
 		return
 	}
+	var quota int64
 	completionRatio := billingratio.GetCompletionRatio(textRequest.Model, meta.ChannelType)
 	promptTokens := usage.PromptTokens
-	// It appears that DeepSeek's official service automatically merges ReasoningTokens into CompletionTokens,
-	// but the behavior of third-party providers may differ, so for now we do not add them manually.
-	// completionTokens := usage.CompletionTokens + usage.CompletionTokensDetails.ReasoningTokens
 	completionTokens := usage.CompletionTokens
-	quota = int64(math.Ceil((float64(promptTokens)+float64(completionTokens)*completionRatio)*ratio)) + usage.ToolsCost
+	quota = int64(math.Ceil((float64(promptTokens) + float64(completionTokens)*completionRatio) * ratio))
 	if ratio != 0 && quota <= 0 {
 		quota = 1
 	}
-
 	totalTokens := promptTokens + completionTokens
 	if totalTokens == 0 {
 		// in this case, must be some error happened
@@ -131,13 +122,7 @@ func postConsumeQuota(ctx context.Context,
 	if err != nil {
 		logger.Error(ctx, "error update user quota cache: "+err.Error())
 	}
-
-	var logContent string
-	if usage.ToolsCost == 0 {
-		logContent = fmt.Sprintf("倍率：%.2f × %.2f × %.2f", modelRatio, groupRatio, completionRatio)
-	} else {
-		logContent = fmt.Sprintf("倍率：%.2f × %.2f × %.2f, tools cost %d", modelRatio, groupRatio, completionRatio, usage.ToolsCost)
-	}
+	logContent := fmt.Sprintf("倍率：%.2f × %.2f × %.2f", modelRatio, groupRatio, completionRatio)
 	model.RecordConsumeLog(ctx, &model.Log{
 		UserId:            meta.UserId,
 		ChannelId:         meta.ChannelId,
@@ -153,8 +138,6 @@ func postConsumeQuota(ctx context.Context,
 	})
 	model.UpdateUserUsedQuotaAndRequestCount(meta.UserId, quota)
 	model.UpdateChannelUsedQuota(meta.ChannelId, quota)
-
-	return quota
 }

 func getMappedModelName(modelName string, mapping map[string]string) (string, bool) {
--- a/relay/controller/text.go
+++ b/relay/controller/text.go
@@ -10,7 +10,6 @@ import (
 	"github.com/gin-gonic/gin"

 	"github.com/songquanpeng/one-api/common/config"
-	"github.com/songquanpeng/one-api/common/ctxkey"
 	"github.com/songquanpeng/one-api/common/logger"
 	"github.com/songquanpeng/one-api/relay"
 	"github.com/songquanpeng/one-api/relay/adaptor"
@@ -105,8 +104,6 @@ func getRequestBody(c *gin.Context, meta *meta.Meta, textRequest *model.GeneralO
 		logger.Debugf(c.Request.Context(), "converted request failed: %s\n", err.Error())
 		return nil, err
 	}
-	c.Set(ctxkey.ConvertedRequest, convertedRequest)
-
 	jsonData, err := json.Marshal(convertedRequest)
 	if err != nil {
 		logger.Debugf(c.Request.Context(), "converted request json_marshal_failed: %s\n", err.Error())
--- a/relay/model/general.go
+++ b/relay/model/general.go
@@ -1,7 +1,5 @@
 package model

-import "github.com/songquanpeng/one-api/relay/adaptor/openrouter"
-
 type ResponseFormat struct {
 	Type       string      `json:"type,omitempty"`
 	JsonSchema *JSONSchema `json:"json_schema,omitempty"`
@@ -25,103 +23,49 @@ type StreamOptions struct {

 type GeneralOpenAIRequest struct {
 	// https://platform.openai.com/docs/api-reference/chat/create
-	Messages []Message `json:"messages,omitempty"`
-	Model    string    `json:"model,omitempty"`
-	Store    *bool     `json:"store,omitempty"`
-	Metadata any       `json:"metadata,omitempty"`
-	// FrequencyPenalty is a number between -2.0 and 2.0 that penalizes
-	// new tokens based on their existing frequency in the text so far,
-	// default is 0.
-	FrequencyPenalty    *float64 `json:"frequency_penalty,omitempty" binding:"omitempty,min=-2,max=2"`
-	LogitBias           any      `json:"logit_bias,omitempty"`
-	Logprobs            *bool    `json:"logprobs,omitempty"`
-	TopLogprobs         *int     `json:"top_logprobs,omitempty"`
-	MaxTokens           int      `json:"max_tokens,omitempty"`
-	MaxCompletionTokens *int     `json:"max_completion_tokens,omitempty"`
-	// N is how many chat completion choices to generate for each input message,
-	// default to 1.
-	N *int `json:"n,omitempty" binding:"omitempty,min=1"`
-	// ReasoningEffort constrains effort on reasoning for reasoning models, reasoning models only.
-	ReasoningEffort *string `json:"reasoning_effort,omitempty" binding:"omitempty,oneof=low medium high"`
-	// Modalities currently the model only programmatically allows modalities = [“text”, “audio”]
-	Modalities []string `json:"modalities,omitempty"`
-	Prediction any      `json:"prediction,omitempty"`
-	Audio      *Audio   `json:"audio,omitempty"`
-	// PresencePenalty is a number between -2.0 and 2.0 that penalizes
-	// new tokens based on whether they appear in the text so far, default is 0.
-	PresencePenalty  *float64        `json:"presence_penalty,omitempty" binding:"omitempty,min=-2,max=2"`
-	ResponseFormat   *ResponseFormat `json:"response_format,omitempty"`
-	Seed             float64         `json:"seed,omitempty"`
-	ServiceTier      *string         `json:"service_tier,omitempty" binding:"omitempty,oneof=default auto"`
-	Stop             any             `json:"stop,omitempty"`
-	Stream           bool            `json:"stream,omitempty"`
-	StreamOptions    *StreamOptions  `json:"stream_options,omitempty"`
-	Temperature      *float64        `json:"temperature,omitempty"`
-	TopP             *float64        `json:"top_p,omitempty"`
-	TopK             int             `json:"top_k,omitempty"`
-	Tools            []Tool          `json:"tools,omitempty"`
-	ToolChoice       any             `json:"tool_choice,omitempty"`
-	ParallelTooCalls *bool           `json:"parallel_tool_calls,omitempty"`
-	User             string          `json:"user,omitempty"`
-	FunctionCall     any             `json:"function_call,omitempty"`
-	Functions        any             `json:"functions,omitempty"`
+	Messages            []Message       `json:"messages,omitempty"`
+	Model               string          `json:"model,omitempty"`
+	Store               *bool           `json:"store,omitempty"`
+	ReasoningEffort     *string         `json:"reasoning_effort,omitempty"`
+	Metadata            any             `json:"metadata,omitempty"`
+	FrequencyPenalty    *float64        `json:"frequency_penalty,omitempty"`
+	LogitBias           any             `json:"logit_bias,omitempty"`
+	Logprobs            *bool           `json:"logprobs,omitempty"`
+	TopLogprobs         *int            `json:"top_logprobs,omitempty"`
+	MaxTokens           int             `json:"max_tokens,omitempty"`
+	MaxCompletionTokens *int            `json:"max_completion_tokens,omitempty"`
+	N                   int             `json:"n,omitempty"`
+	Modalities          []string        `json:"modalities,omitempty"`
+	Prediction          any             `json:"prediction,omitempty"`
+	Audio               *Audio          `json:"audio,omitempty"`
+	PresencePenalty     *float64        `json:"presence_penalty,omitempty"`
+	ResponseFormat      *ResponseFormat `json:"response_format,omitempty"`
+	Seed                float64         `json:"seed,omitempty"`
+	ServiceTier         *string         `json:"service_tier,omitempty"`
+	Stop                any             `json:"stop,omitempty"`
+	Stream              bool            `json:"stream,omitempty"`
+	StreamOptions       *StreamOptions  `json:"stream_options,omitempty"`
+	Temperature         *float64        `json:"temperature,omitempty"`
+	TopP                *float64        `json:"top_p,omitempty"`
+	TopK                int             `json:"top_k,omitempty"`
+	Tools               []Tool          `json:"tools,omitempty"`
+	ToolChoice          any             `json:"tool_choice,omitempty"`
+	ParallelTooCalls    *bool           `json:"parallel_tool_calls,omitempty"`
+	User                string          `json:"user,omitempty"`
+	FunctionCall        any             `json:"function_call,omitempty"`
+	Functions           any             `json:"functions,omitempty"`
 	// https://platform.openai.com/docs/api-reference/embeddings/create
 	Input          any    `json:"input,omitempty"`
 	EncodingFormat string `json:"encoding_format,omitempty"`
 	Dimensions     int    `json:"dimensions,omitempty"`
 	// https://platform.openai.com/docs/api-reference/images/create
-	Prompt           string            `json:"prompt,omitempty"`
-	Quality          *string           `json:"quality,omitempty"`
-	Size             string            `json:"size,omitempty"`
-	Style            *string           `json:"style,omitempty"`
-	WebSearchOptions *WebSearchOptions `json:"web_search_options,omitempty"`
-
+	Prompt  any     `json:"prompt,omitempty"`
+	Quality *string `json:"quality,omitempty"`
+	Size    string  `json:"size,omitempty"`
+	Style   *string `json:"style,omitempty"`
 	// Others
 	Instruction string `json:"instruction,omitempty"`
 	NumCtx      int    `json:"num_ctx,omitempty"`
-	// -------------------------------------
-	// Openrouter
-	// -------------------------------------
-	Provider         *openrouter.RequestProvider `json:"provider,omitempty"`
-	IncludeReasoning *bool                       `json:"include_reasoning,omitempty"`
-	// -------------------------------------
-	// Anthropic
-	// -------------------------------------
-	Thinking *Thinking `json:"thinking,omitempty"`
-}
-
-// WebSearchOptions is the tool searches the web for relevant results to use in a response.
-type WebSearchOptions struct {
-	// SearchContextSize is the high level guidance for the amount of context window space to use for the search,
-	// default is "medium".
-	SearchContextSize *string       `json:"search_context_size,omitempty" binding:"omitempty,oneof=low medium high"`
-	UserLocation      *UserLocation `json:"user_location,omitempty"`
-}
-
-// UserLocation is a struct that contains the location of the user.
-type UserLocation struct {
-	// Approximate is the approximate location parameters for the search.
-	Approximate UserLocationApproximate `json:"approximate" binding:"required"`
-	// Type is the type of location approximation.
-	Type string `json:"type" binding:"required,oneof=approximate"`
-}
-
-// UserLocationApproximate is a struct that contains the approximate location of the user.
-type UserLocationApproximate struct {
-	// City is the city of the user, e.g. San Francisco.
-	City *string `json:"city,omitempty"`
-	// Country is the country of the user, e.g. US.
-	Country *string `json:"country,omitempty"`
-	// Region is the region of the user, e.g. California.
-	Region *string `json:"region,omitempty"`
-	// Timezone is the IANA timezone of the user, e.g. America/Los_Angeles.
-	Timezone *string `json:"timezone,omitempty"`
-}
-
-// https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#implementing-extended-thinking
-type Thinking struct {
-	Type         string `json:"type"`
-	BudgetTokens int    `json:"budget_tokens" binding:"omitempty,min=1024"`
 }

 func (r GeneralOpenAIRequest) ParseInput() []string {
--- a/relay/model/message.go
+++ b/relay/model/message.go
@@ -1,106 +1,12 @@
 package model

-import (
-	"context"
-	"strings"
-
-	"github.com/songquanpeng/one-api/common/logger"
-)
-
-// ReasoningFormat is the format of reasoning content,
-// can be set by the reasoning_format parameter in the request url.
-type ReasoningFormat string
-
-const (
-	ReasoningFormatUnspecified ReasoningFormat = ""
-	// ReasoningFormatReasoningContent is the reasoning format used by deepseek official API
-	ReasoningFormatReasoningContent ReasoningFormat = "reasoning_content"
-	// ReasoningFormatReasoning is the reasoning format used by openrouter
-	ReasoningFormatReasoning ReasoningFormat = "reasoning"
-
-	// ReasoningFormatThinkTag is the reasoning format used by 3rd party deepseek-r1 providers.
-	//
-	// Deprecated: I believe <think> is a very poor format, especially in stream mode, it is difficult to extract and convert.
-	// Considering that only a few deepseek-r1 third-party providers use this format, it has been decided to no longer support it.
-	// ReasoningFormatThinkTag ReasoningFormat = "think-tag"
-
-	// ReasoningFormatThinking is the reasoning format used by anthropic
-	ReasoningFormatThinking ReasoningFormat = "thinking"
-)
-
 type Message struct {
-	Role string `json:"role,omitempty"`
-	// Content is a string or a list of objects
-	Content    any              `json:"content,omitempty"`
-	Name       *string          `json:"name,omitempty"`
-	ToolCalls  []Tool           `json:"tool_calls,omitempty"`
-	ToolCallId string           `json:"tool_call_id,omitempty"`
-	Audio      *messageAudio    `json:"audio,omitempty"`
-	Annotation []AnnotationItem `json:"annotation,omitempty"`
-
-	// -------------------------------------
-	// Deepseek 专有的一些字段
-	// https://api-docs.deepseek.com/api/create-chat-completion
-	// -------------------------------------
-	// Prefix forces the model to begin its answer with the supplied prefix in the assistant message.
-	// To enable this feature, set base_url to "https://api.deepseek.com/beta".
-	Prefix *bool `json:"prefix,omitempty"` // ReasoningContent is Used for the deepseek-reasoner model in the Chat
-	// Prefix Completion feature as the input for the CoT in the last assistant message.
-	// When using this feature, the prefix parameter must be set to true.
-	ReasoningContent *string `json:"reasoning_content,omitempty"`
-
-	// -------------------------------------
-	// Openrouter
-	// -------------------------------------
-	Reasoning *string `json:"reasoning,omitempty"`
-	Refusal   *bool   `json:"refusal,omitempty"`
-
-	// -------------------------------------
-	// Anthropic
-	// -------------------------------------
-	Thinking  *string `json:"thinking,omitempty"`
-	Signature *string `json:"signature,omitempty"`
-}
-
-type AnnotationItem struct {
-	Type        string      `json:"type" binding:"oneof=url_citation"`
-	UrlCitation UrlCitation `json:"url_citation"`
-}
-
-// UrlCitation is a URL citation when using web search.
-type UrlCitation struct {
-	// Endpoint is the index of the last character of the URL citation in the message.
-	EndIndex int `json:"end_index"`
-	// StartIndex is the index of the first character of the URL citation in the message.
-	StartIndex int `json:"start_index"`
-	// Title is the title of the web resource.
-	Title string `json:"title"`
-	// Url is the URL of the web resource.
-	Url string `json:"url"`
-}
-
-// SetReasoningContent sets the reasoning content based on the format
-func (m *Message) SetReasoningContent(format string, reasoningContent string) {
-	switch ReasoningFormat(strings.ToLower(strings.TrimSpace(format))) {
-	case ReasoningFormatReasoningContent:
-		m.ReasoningContent = &reasoningContent
-		// case ReasoningFormatThinkTag:
-		// 	m.Content = fmt.Sprintf("<think>%s</think>%s", reasoningContent, m.Content)
-	case ReasoningFormatThinking:
-		m.Thinking = &reasoningContent
-	case ReasoningFormatReasoning,
-		ReasoningFormatUnspecified:
-		m.Reasoning = &reasoningContent
-	default:
-		logger.Warnf(context.TODO(), "unknown reasoning format: %q", format)
-	}
-}
-
-type messageAudio struct {
-	Id         string `json:"id"`
-	Data       string `json:"data,omitempty"`
-	ExpiredAt  int    `json:"expired_at,omitempty"`
-	Transcript string `json:"transcript,omitempty"`
+	Role             string  `json:"role,omitempty"`
+	Content          any     `json:"content,omitempty"`
+	ReasoningContent any     `json:"reasoning_content,omitempty"`
+	Name             *string `json:"name,omitempty"`
+	ToolCalls        []Tool  `json:"tool_calls,omitempty"`
+	ToolCallId       string  `json:"tool_call_id,omitempty"`
 }

 func (m Message) IsStringContent() bool {
@@ -121,7 +27,6 @@ func (m Message) StringContent() string {
 			if !ok {
 				continue
 			}
-
 			if contentMap["type"] == ContentTypeText {
 				if subStr, ok := contentMap["text"].(string); ok {
 					contentStr += subStr
@@ -130,7 +35,6 @@ func (m Message) StringContent() string {
 		}
 		return contentStr
 	}
-
 	return ""
 }

@@ -144,7 +48,6 @@ func (m Message) ParseContent() []MessageContent {
 		})
 		return contentList
 	}
-
 	anyList, ok := m.Content.([]any)
 	if ok {
 		for _, contentItem := range anyList {
@@ -169,21 +72,8 @@ func (m Message) ParseContent() []MessageContent {
 						},
 					})
 				}
-			case ContentTypeInputAudio:
-				if subObj, ok := contentMap["input_audio"].(map[string]any); ok {
-					contentList = append(contentList, MessageContent{
-						Type: ContentTypeInputAudio,
-						InputAudio: &InputAudio{
-							Data:   subObj["data"].(string),
-							Format: subObj["format"].(string),
-						},
-					})
-				}
-			default:
-				logger.Warnf(context.TODO(), "unknown content type: %s", contentMap["type"])
 			}
 		}
-
 		return contentList
 	}
 	return nil
@@ -195,23 +85,7 @@ type ImageURL struct {
 }

 type MessageContent struct {
-	// Type should be one of the following: text/input_audio
-	Type       string      `json:"type,omitempty"`
-	Text       string      `json:"text"`
-	ImageURL   *ImageURL   `json:"image_url,omitempty"`
-	InputAudio *InputAudio `json:"input_audio,omitempty"`
-	// -------------------------------------
-	// Anthropic
-	// -------------------------------------
-	Thinking  *string `json:"thinking,omitempty"`
-	Signature *string `json:"signature,omitempty"`
-}
-
-type InputAudio struct {
-	// Data is the base64 encoded audio data
-	Data string `json:"data" binding:"required"`
-	// Format is the audio format, should be one of the
-	// following: mp3/mp4/mpeg/mpga/m4a/wav/webm/pcm16.
-	// When stream=true, format should be pcm16
-	Format string `json:"format"`
+	Type     string    `json:"type,omitempty"`
+	Text     string    `json:"text"`
+	ImageURL *ImageURL `json:"image_url,omitempty"`
 }
--- a/relay/model/misc.go
+++ b/relay/model/misc.go
@@ -1,22 +1,17 @@
 package model

-// Usage is the token usage information returned by OpenAI API.
 type Usage struct {
 	PromptTokens     int `json:"prompt_tokens"`
 	CompletionTokens int `json:"completion_tokens"`
 	TotalTokens      int `json:"total_tokens"`
-	// PromptTokensDetails may be empty for some models
-	PromptTokensDetails *usagePromptTokensDetails `json:"prompt_tokens_details,omitempty"`
-	// CompletionTokensDetails may be empty for some models
-	CompletionTokensDetails *usageCompletionTokensDetails `json:"completion_tokens_details,omitempty"`
-	ServiceTier             string                        `json:"service_tier,omitempty"`
-	SystemFingerprint       string                        `json:"system_fingerprint,omitempty"`

-	// -------------------------------------
-	// Custom fields
-	// -------------------------------------
-	// ToolsCost is the cost of using tools, in quota.
-	ToolsCost int64 `json:"tools_cost,omitempty"`
+	CompletionTokensDetails *CompletionTokensDetails `json:"completion_tokens_details,omitempty"`
+}
+
+type CompletionTokensDetails struct {
+	ReasoningTokens          int `json:"reasoning_tokens"`
+	AcceptedPredictionTokens int `json:"accepted_prediction_tokens"`
+	RejectedPredictionTokens int `json:"rejected_prediction_tokens"`
 }

 type Error struct {
@@ -30,20 +25,3 @@ type ErrorWithStatusCode struct {
 	Error
 	StatusCode int `json:"status_code"`
 }
-
-type usagePromptTokensDetails struct {
-	CachedTokens int `json:"cached_tokens"`
-	AudioTokens  int `json:"audio_tokens"`
-	// TextTokens could be zero for pure text chats
-	TextTokens  int `json:"text_tokens"`
-	ImageTokens int `json:"image_tokens"`
-}
-
-type usageCompletionTokensDetails struct {
-	ReasoningTokens          int `json:"reasoning_tokens"`
-	AudioTokens              int `json:"audio_tokens"`
-	AcceptedPredictionTokens int `json:"accepted_prediction_tokens"`
-	RejectedPredictionTokens int `json:"rejected_prediction_tokens"`
-	// TextTokens could be zero for pure text chats
-	TextTokens int `json:"text_tokens"`
-}