feat: support Responses to Chat (#5787)

* fix(openai): harden Chat-to-Responses compatibility

Add a shared Responses-to-Chat stream state machine and use it from the OpenAI relay path. Preserve assistant text alongside tool calls, bind tool argument deltas by output_index, map incomplete finish reasons, support reasoning/custom tool events, and buffer upstream SSE for non-stream Chat clients.

Add deterministic service tests and relay SSE tests for the conversion path.

Related to #5745.

* refactor: rename openaicompat to relayconvert for improved clarity

* feat(gemini): support responses request conversion

* feat: add responses to chat conversion support

* fix: harden responses chat conversion edge cases

.gitignore

@@ -39,4 +39,5 @@ skills-lock.json
 
 # Local-only live probes and scratch test workspaces.
 .local-tests/
+service/relayconvert/chat_responses_live_local_test.go
 service/openaicompat/chat_responses_live_local_test.go

dto/channel_settings.go

@@ -63,6 +63,7 @@ const (
 	AdvancedCustomConverterAnthropicMessagesToOpenAIChatCompletions     = "anthropic_messages_to_openai_chat_completions"
 	AdvancedCustomConverterOpenAIChatCompletionsToAnthropicMessages     = "openai_chat_completions_to_anthropic_messages"
 	AdvancedCustomConverterOpenAIChatCompletionsToOpenAIResponses       = "openai_chat_completions_to_openai_responses"
+	AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions       = "openai_responses_to_openai_chat_completions"
 	AdvancedCustomConverterGeminiGenerateContentToOpenAIChatCompletions = "gemini_generate_content_to_openai_chat_completions"
 	AdvancedCustomConverterOpenAIChatCompletionsToGeminiGenerateContent = "openai_chat_completions_to_gemini_generate_content"
 )
@@ -147,6 +148,7 @@ func IsAdvancedCustomConverterAllowed(converter string) bool {
 		AdvancedCustomConverterAnthropicMessagesToOpenAIChatCompletions,
 		AdvancedCustomConverterOpenAIChatCompletionsToAnthropicMessages,
 		AdvancedCustomConverterOpenAIChatCompletionsToOpenAIResponses,
+		AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions,
 		AdvancedCustomConverterGeminiGenerateContentToOpenAIChatCompletions,
 		AdvancedCustomConverterOpenAIChatCompletionsToGeminiGenerateContent:
 		return true
@@ -240,6 +242,10 @@ func validateAdvancedCustomConverterPath(index int, incomingPath string, convert
 		if incomingPath == "/v1/chat/completions" {
 			return nil
 		}
+	case AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions:
+		if incomingPath == "/v1/responses" {
+			return nil
+		}
 	case AdvancedCustomConverterGeminiGenerateContentToOpenAIChatCompletions:
 		if strings.Contains(incomingPath, ":generateContent") || strings.Contains(incomingPath, ":streamGenerateContent") {
 			return nil

dto/channel_settings_test.go

+package dto
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestAdvancedCustomValidateResponsesToChatConverterPath(t *testing.T) {
+	valid := &AdvancedCustomConfig{
+		Routes: []AdvancedCustomRoute{
+			{
+				IncomingPath: "/v1/responses",
+				UpstreamPath: "/v1/chat/completions",
+				Converter:    AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions,
+			},
+		},
+	}
+	require.NoError(t, valid.Validate())
+
+	tests := []struct {
+		name         string
+		incomingPath string
+	}{
+		{name: "chat completions", incomingPath: "/v1/chat/completions"},
+		{name: "responses compact", incomingPath: "/v1/responses/compact"},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			config := &AdvancedCustomConfig{
+				Routes: []AdvancedCustomRoute{
+					{
+						IncomingPath: tt.incomingPath,
+						UpstreamPath: "/v1/chat/completions",
+						Converter:    AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions,
+					},
+				},
+			}
+			err := config.Validate()
+			require.Error(t, err)
+			assert.Contains(t, err.Error(), "converter does not match incoming_path")
+		})
+	}
+}

relay/channel/advancedcustom/adaptor.go

@@ -104,10 +104,18 @@ func (a *Adaptor) ConvertOpenAIResponsesRequest(c *gin.Context, info *relaycommo
 	if err != nil {
 		return nil, err
 	}
-	if converter != dto.AdvancedCustomConverterNone {
+	switch converter {
+	case dto.AdvancedCustomConverterNone:
+		return a.convertOpenAICompatibleResponsesRequest(c, info, request)
+	case dto.AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions:
+		chatReq, err := service.ResponsesRequestToChatCompletionsRequest(&request)
+		if err != nil {
+			return nil, err
+		}
+		return a.convertOpenAICompatibleRequest(c, info, chatReq)
+	default:
 		return nil, fmt.Errorf("converter %q does not support OpenAI Responses requests", converter)
 	}
-	return a.convertOpenAICompatibleResponsesRequest(c, info, request)
 }
 
 func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.EmbeddingRequest) (any, error) {
@@ -221,6 +229,11 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 			return openai.OaiResponsesToChatStreamHandler(c, info, resp)
 		}
 		return openai.OaiResponsesToChatHandler(c, info, resp)
+	case dto.AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions:
+		if info.IsStream {
+			return openai.OaiChatToResponsesStreamHandler(c, info, resp)
+		}
+		return openai.OaiChatToResponsesHandler(c, info, resp)
 	default:
 		return nil, types.NewOpenAIError(fmt.Errorf("unsupported advanced custom converter: %s", a.converter), types.ErrorCodeInvalidRequest, http.StatusBadRequest, types.ErrOptionWithSkipRetry())
 	}

relay/channel/advancedcustom/adaptor_test.go

@@ -6,6 +6,7 @@ import (
 	"net/url"
 	"testing"
 
+	"github.com/QuantumNous/new-api/common"
 	"github.com/QuantumNous/new-api/constant"
 	"github.com/QuantumNous/new-api/dto"
 	relaycommon "github.com/QuantumNous/new-api/relay/common"
@@ -279,6 +280,44 @@ func TestAdaptorMatchesGeminiIncomingPathTemplate(t *testing.T) {
 	}
 }
 
+func TestAdaptorConvertsResponsesRequestToOpenAIChatUpstream(t *testing.T) {
+	adaptor := &Adaptor{}
+	info := advancedCustomRelayInfo(&dto.AdvancedCustomConfig{
+		Routes: []dto.AdvancedCustomRoute{
+			{
+				IncomingPath: "/v1/responses",
+				UpstreamPath: "/v1/chat/completions",
+				Converter:    dto.AdvancedCustomConverterOpenAIResponsesToOpenAIChatCompletions,
+			},
+		},
+	})
+	info.RelayMode = relayconstant.RelayModeResponses
+	info.RequestURLPath = "/v1/responses"
+	c := advancedCustomGinContext("/v1/responses")
+
+	converted, err := adaptor.ConvertOpenAIResponsesRequest(c, info, dto.OpenAIResponsesRequest{
+		Model:        "gpt-test",
+		Instructions: mustAdvancedCustomRawMessage(t, "system rules"),
+		Input:        mustAdvancedCustomRawMessage(t, "hello"),
+	})
+	require.NoError(t, err)
+
+	chatReq, ok := converted.(*dto.GeneralOpenAIRequest)
+	require.True(t, ok)
+	assert.Equal(t, "gpt-test", chatReq.Model)
+	require.Len(t, chatReq.Messages, 2)
+	assert.Equal(t, "system", chatReq.Messages[0].Role)
+	assert.Equal(t, "system rules", chatReq.Messages[0].StringContent())
+	assert.Equal(t, "user", chatReq.Messages[1].Role)
+	assert.Equal(t, "hello", chatReq.Messages[1].StringContent())
+
+	requestURL, err := adaptor.GetRequestURL(info)
+	require.NoError(t, err)
+	parsedURL, err := url.Parse(requestURL)
+	require.NoError(t, err)
+	assert.Equal(t, "/v1/chat/completions", parsedURL.Path)
+}
+
 func advancedCustomRelayInfo(config *dto.AdvancedCustomConfig) *relaycommon.RelayInfo {
 	return &relaycommon.RelayInfo{
 		RelayFormat:    types.RelayFormatOpenAI,
@@ -302,3 +341,10 @@ func advancedCustomGinContext(path string) *gin.Context {
 	c.Request.Header.Set("Content-Type", "application/json")
 	return c
 }
+
+func mustAdvancedCustomRawMessage(t *testing.T, value any) []byte {
+	t.Helper()
+	raw, err := common.Marshal(value)
+	require.NoError(t, err)
+	return raw
+}

relay/channel/claude/relay_claude_test.go

@@ -5,6 +5,8 @@ import (
 	"testing"
 
 	"github.com/QuantumNous/new-api/dto"
+	"github.com/QuantumNous/new-api/service"
+	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
@@ -12,6 +14,48 @@ func commonPointer[T any](value T) *T {
 	return &value
 }
 
+func TestResponseOpenAI2ClaudeToolUseInputIsObject(t *testing.T) {
+	tests := []struct {
+		name string
+		args string
+		want map[string]interface{}
+	}{
+		{name: "object", args: `{"q":"x"}`, want: map[string]interface{}{"q": "x"}},
+		{name: "empty", args: "", want: map[string]interface{}{}},
+		{name: "invalid", args: "{", want: map[string]interface{}{}},
+		{name: "null", args: "null", want: map[string]interface{}{}},
+		{name: "array", args: `["x"]`, want: map[string]interface{}{}},
+		{name: "string", args: `"x"`, want: map[string]interface{}{}},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			msg := dto.Message{Role: "assistant"}
+			msg.SetToolCalls([]dto.ToolCallRequest{
+				{
+					ID:   "call_1",
+					Type: "function",
+					Function: dto.FunctionRequest{
+						Name:      "lookup",
+						Arguments: tt.args,
+					},
+				},
+			})
+			resp := service.ResponseOpenAI2Claude(&dto.OpenAITextResponse{
+				Id:    "chatcmpl_1",
+				Model: "gpt-test",
+				Choices: []dto.OpenAITextResponseChoice{
+					{Message: msg, FinishReason: "tool_calls"},
+				},
+			}, nil)
+
+			require.Len(t, resp.Content, 1)
+			assert.Equal(t, "tool_use", resp.Content[0].Type)
+			assert.Equal(t, tt.want, resp.Content[0].Input)
+		})
+	}
+}
+
 func TestFormatClaudeResponseInfo_MessageStart(t *testing.T) {
 	claudeInfo := &ClaudeResponseInfo{
 		Usage: &dto.Usage{},

relay/channel/gemini/adaptor.go

@@ -12,6 +12,7 @@ import (
 	"github.com/QuantumNous/new-api/relay/channel/openai"
 	relaycommon "github.com/QuantumNous/new-api/relay/common"
 	"github.com/QuantumNous/new-api/relay/constant"
+	"github.com/QuantumNous/new-api/service/relayconvert"
 	"github.com/QuantumNous/new-api/setting/model_setting"
 	"github.com/QuantumNous/new-api/setting/reasoning"
 	"github.com/QuantumNous/new-api/types"
@@ -238,8 +239,17 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 }
 
 func (a *Adaptor) ConvertOpenAIResponsesRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.OpenAIResponsesRequest) (any, error) {
-	// TODO implement me
-	return nil, errors.New("not implemented")
+	request, err := preprocessGeminiOpenAIResponsesRequest(request)
+	if err != nil {
+		return nil, err
+	}
+
+	chatRequest, err := relayconvert.ResponsesRequestToChatCompletionsRequest(&request)
+	if err != nil {
+		return nil, err
+	}
+
+	return a.ConvertOpenAIRequest(c, info, chatRequest)
 }
 
 func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (any, error) {
@@ -247,6 +257,13 @@ 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.RelayMode == constant.RelayModeResponses {
+		if info.IsStream {
+			return GeminiResponsesStreamHandler(c, info, resp)
+		}
+		return GeminiResponsesHandler(c, info, resp)
+	}
+
 	if info.RelayMode == constant.RelayModeGemini {
 		if strings.Contains(info.RequestURLPath, ":embedContent") ||
 			strings.Contains(info.RequestURLPath, ":batchEmbedContents") {

relay/channel/gemini/adaptor_responses.go

+package gemini
+
+import (
+	"strings"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/dto"
+)
+
+const (
+	geminiResponsesInputTypeCustomToolCall       = "custom_tool_call"
+	geminiResponsesInputTypeCustomToolCallOutput = "custom_tool_call_output"
+	geminiResponsesInputTypeFunctionCallOutput   = "function_call_output"
+)
+
+func preprocessGeminiOpenAIResponsesRequest(request dto.OpenAIResponsesRequest) (dto.OpenAIResponsesRequest, error) {
+	tools, err := filterGeminiResponsesTools(request.Tools)
+	if err != nil {
+		return request, err
+	}
+	request.Tools = tools
+
+	input, err := filterGeminiResponsesInput(request.Input)
+	if err != nil {
+		return request, err
+	}
+	request.Input = input
+
+	return request, nil
+}
+
+func filterGeminiResponsesTools(raw []byte) ([]byte, error) {
+	if !geminiRawJSONPresent(raw) || common.GetJsonType(raw) != "array" {
+		return raw, nil
+	}
+
+	var tools []map[string]any
+	if err := common.Unmarshal(raw, &tools); err != nil {
+		return nil, err
+	}
+
+	filtered := make([]map[string]any, 0, len(tools))
+	for _, tool := range tools {
+		if strings.TrimSpace(common.Interface2String(tool["type"])) != "function" {
+			// TODO: Support Responses custom/freeform tools when Gemini has a safe equivalent representation.
+			continue
+		}
+		filtered = append(filtered, tool)
+	}
+	if len(filtered) == 0 {
+		return nil, nil
+	}
+	return common.Marshal(filtered)
+}
+
+func filterGeminiResponsesInput(raw []byte) ([]byte, error) {
+	if !geminiRawJSONPresent(raw) || common.GetJsonType(raw) != "array" {
+		return raw, nil
+	}
+
+	var items []map[string]any
+	if err := common.Unmarshal(raw, &items); err != nil {
+		return nil, err
+	}
+
+	skippedCustomCallIDs := make(map[string]struct{})
+	for _, item := range items {
+		if strings.TrimSpace(common.Interface2String(item["type"])) != geminiResponsesInputTypeCustomToolCall {
+			continue
+		}
+		if callID := strings.TrimSpace(common.Interface2String(item["call_id"])); callID != "" {
+			skippedCustomCallIDs[callID] = struct{}{}
+		}
+	}
+
+	filtered := make([]map[string]any, 0, len(items))
+	for _, item := range items {
+		itemType := strings.TrimSpace(common.Interface2String(item["type"]))
+		switch itemType {
+		case geminiResponsesInputTypeCustomToolCall, geminiResponsesInputTypeCustomToolCallOutput:
+			// TODO: Support Responses custom/freeform tool calls once Gemini can preserve their semantics.
+			continue
+		case geminiResponsesInputTypeFunctionCallOutput:
+			if _, ok := skippedCustomCallIDs[strings.TrimSpace(common.Interface2String(item["call_id"]))]; ok {
+				continue
+			}
+		}
+		filtered = append(filtered, item)
+	}
+
+	return common.Marshal(filtered)
+}
+
+func geminiRawJSONPresent(raw []byte) bool {
+	if len(raw) == 0 {
+		return false
+	}
+	return common.GetJsonType(raw) != "null"
+}

relay/channel/gemini/adaptor_responses_test.go

+package gemini
+
+import (
+	"testing"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/dto"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"github.com/tidwall/gjson"
+)
+
+func TestConvertOpenAIResponsesRequestToGeminiInstructionsAndInput(t *testing.T) {
+	got := mustConvertResponsesToGemini(t, dto.OpenAIResponsesRequest{
+		Model:        "gemini-test",
+		Instructions: mustGeminiRawMessage(t, "system rules"),
+		Input:        mustGeminiRawMessage(t, "hello"),
+	})
+
+	require.NotNil(t, got.SystemInstructions)
+	require.Len(t, got.SystemInstructions.Parts, 1)
+	assert.Equal(t, "system rules", got.SystemInstructions.Parts[0].Text)
+	require.Len(t, got.Contents, 1)
+	assert.Equal(t, "user", got.Contents[0].Role)
+	require.Len(t, got.Contents[0].Parts, 1)
+	assert.Equal(t, "hello", got.Contents[0].Parts[0].Text)
+}
+
+func TestConvertOpenAIResponsesRequestToGeminiFunctionToolAndChoice(t *testing.T) {
+	got := mustConvertResponsesToGemini(t, dto.OpenAIResponsesRequest{
+		Model: "gemini-test",
+		Input: mustGeminiRawMessage(t, "lookup weather"),
+		Tools: mustGeminiRawMessage(t, []map[string]any{
+			{
+				"type":        "function",
+				"name":        "lookup",
+				"description": "Lookup data",
+				"parameters": map[string]any{
+					"type": "object",
+					"properties": map[string]any{
+						"q": map[string]any{"type": "string"},
+					},
+				},
+			},
+			{"type": "custom", "name": "freeform"},
+		}),
+		ToolChoice: mustGeminiRawMessage(t, map[string]any{
+			"type": "function",
+			"name": "lookup",
+		}),
+	})
+
+	tools := got.GetTools()
+	require.Len(t, tools, 1)
+	assert.Equal(t, "lookup", gjson.GetBytes(got.Tools, "0.functionDeclarations.0.name").String())
+	assert.Equal(t, "Lookup data", gjson.GetBytes(got.Tools, "0.functionDeclarations.0.description").String())
+	require.NotNil(t, got.ToolConfig)
+	require.NotNil(t, got.ToolConfig.FunctionCallingConfig)
+	assert.Equal(t, dto.FunctionCallingConfigMode("ANY"), got.ToolConfig.FunctionCallingConfig.Mode)
+	assert.Equal(t, []string{"lookup"}, got.ToolConfig.FunctionCallingConfig.AllowedFunctionNames)
+}
+
+func TestConvertOpenAIResponsesRequestToGeminiFunctionCallConversation(t *testing.T) {
+	got := mustConvertResponsesToGemini(t, dto.OpenAIResponsesRequest{
+		Model: "gemini-test",
+		Input: mustGeminiRawMessage(t, []map[string]any{
+			{
+				"role": "assistant",
+				"content": []map[string]any{
+					{"type": "output_text", "text": "I will call."},
+				},
+			},
+			{
+				"type":      "function_call",
+				"call_id":   "call_1",
+				"name":      "lookup",
+				"arguments": map[string]any{"q": "x"},
+			},
+			{
+				"type":    "function_call_output",
+				"call_id": "call_1",
+				"output":  map[string]any{"ok": true},
+			},
+		}),
+		Tools: mustGeminiRawMessage(t, []map[string]any{
+			{"type": "function", "name": "lookup", "parameters": map[string]any{"type": "object"}},
+		}),
+	})
+
+	require.Len(t, got.Contents, 2)
+	assert.Equal(t, "model", got.Contents[0].Role)
+	require.Len(t, got.Contents[0].Parts, 2)
+	require.NotNil(t, got.Contents[0].Parts[0].FunctionCall)
+	assert.Equal(t, "lookup", got.Contents[0].Parts[0].FunctionCall.FunctionName)
+	assert.Equal(t, map[string]interface{}{"q": "x"}, got.Contents[0].Parts[0].FunctionCall.Arguments)
+	assert.Equal(t, "I will call.", got.Contents[0].Parts[1].Text)
+
+	assert.Equal(t, "user", got.Contents[1].Role)
+	require.Len(t, got.Contents[1].Parts, 1)
+	require.NotNil(t, got.Contents[1].Parts[0].FunctionResponse)
+	assert.Equal(t, "lookup", got.Contents[1].Parts[0].FunctionResponse.Name)
+	assert.Equal(t, map[string]interface{}{"ok": true}, got.Contents[1].Parts[0].FunctionResponse.Response)
+}
+
+func TestConvertOpenAIResponsesRequestToGeminiSkipsCustomToolCalls(t *testing.T) {
+	got := mustConvertResponsesToGemini(t, dto.OpenAIResponsesRequest{
+		Model: "gemini-test",
+		Input: mustGeminiRawMessage(t, []map[string]any{
+			{
+				"role": "assistant",
+				"content": []map[string]any{
+					{"type": "output_text", "text": "before custom"},
+				},
+			},
+			{
+				"type":    "custom_tool_call",
+				"call_id": "call_custom",
+				"name":    "apply_patch",
+				"input":   "patch body",
+			},
+			{
+				"type":    "custom_tool_call_output",
+				"call_id": "call_custom",
+				"output":  "ok",
+			},
+			{
+				"type":    "function_call_output",
+				"call_id": "call_custom",
+				"output":  "legacy custom output",
+			},
+			{
+				"role":    "user",
+				"content": "next turn",
+			},
+		}),
+		Tools: mustGeminiRawMessage(t, []map[string]any{
+			{"type": "custom", "name": "apply_patch"},
+			{"type": "unknown", "name": "unknown"},
+		}),
+	})
+
+	assert.Empty(t, got.GetTools())
+	require.Len(t, got.Contents, 2)
+	assert.Equal(t, "model", got.Contents[0].Role)
+	require.Len(t, got.Contents[0].Parts, 1)
+	assert.Equal(t, "before custom", got.Contents[0].Parts[0].Text)
+	assert.Nil(t, got.Contents[0].Parts[0].FunctionCall)
+
+	assert.Equal(t, "user", got.Contents[1].Role)
+	require.Len(t, got.Contents[1].Parts, 1)
+	assert.Equal(t, "next turn", got.Contents[1].Parts[0].Text)
+	assert.Nil(t, got.Contents[1].Parts[0].FunctionResponse)
+}
+
+func mustConvertResponsesToGemini(t *testing.T, req dto.OpenAIResponsesRequest) *dto.GeminiChatRequest {
+	t.Helper()
+	info := &relaycommon.RelayInfo{
+		OriginModelName: req.Model,
+		ChannelMeta: &relaycommon.ChannelMeta{
+			UpstreamModelName: req.Model,
+		},
+	}
+	got, err := (&Adaptor{}).ConvertOpenAIResponsesRequest(nil, info, req)
+	require.NoError(t, err)
+	geminiReq, ok := got.(*dto.GeminiChatRequest)
+	require.True(t, ok)
+	return geminiReq
+}
+
+func mustGeminiRawMessage(t *testing.T, value any) []byte {
+	t.Helper()
+	raw, err := common.Marshal(value)
+	require.NoError(t, err)
+	return raw
+}

relay/channel/gemini/relay_responses.go

+package gemini
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/constant"
+	"github.com/QuantumNous/new-api/dto"
+	"github.com/QuantumNous/new-api/logger"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	"github.com/QuantumNous/new-api/relay/helper"
+	"github.com/QuantumNous/new-api/service"
+	"github.com/QuantumNous/new-api/service/relayconvert"
+	"github.com/QuantumNous/new-api/types"
+	"github.com/gin-gonic/gin"
+)
+
+func GeminiResponsesHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	defer service.CloseResponseBodyGracefully(resp)
+
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponseBody, http.StatusInternalServerError)
+	}
+	logger.LogDebug(c, "Gemini responses response body: %s", responseBody)
+
+	var geminiResponse dto.GeminiChatResponse
+	if err := common.Unmarshal(responseBody, &geminiResponse); err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponseBody, http.StatusInternalServerError)
+	}
+	if len(geminiResponse.Candidates) == 0 {
+		usage := buildUsageFromGeminiMetadata(geminiResponse.UsageMetadata, info.GetEstimatePromptTokens())
+		if geminiResponse.PromptFeedback != nil && geminiResponse.PromptFeedback.BlockReason != nil {
+			common.SetContextKey(c, constant.ContextKeyAdminRejectReason, fmt.Sprintf("gemini_block_reason=%s", *geminiResponse.PromptFeedback.BlockReason))
+			return &usage, types.NewOpenAIError(
+				errors.New("request blocked by Gemini API: "+*geminiResponse.PromptFeedback.BlockReason),
+				types.ErrorCodePromptBlocked,
+				http.StatusBadRequest,
+			)
+		}
+		common.SetContextKey(c, constant.ContextKeyAdminRejectReason, "gemini_empty_candidates")
+		return &usage, types.NewOpenAIError(
+			errors.New("empty response from Gemini API"),
+			types.ErrorCodeEmptyResponse,
+			http.StatusInternalServerError,
+		)
+	}
+
+	chatResp := responseGeminiChat2OpenAI(c, &geminiResponse)
+	chatResp.Model = info.UpstreamModelName
+	usage := buildUsageFromGeminiMetadata(geminiResponse.UsageMetadata, info.GetEstimatePromptTokens())
+	chatResp.Usage = usage
+
+	responsesResp, responsesUsage, err := service.ChatCompletionsResponseToResponsesResponse(chatResp, helper.GetResponseID(c))
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponseBody, http.StatusInternalServerError)
+	}
+	if responsesUsage == nil || responsesUsage.TotalTokens == 0 {
+		responsesResp.Usage = relayconvert.UsageFromChatUsage(&usage)
+	}
+
+	responseBody, err = common.Marshal(responsesResp)
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
+	}
+	service.IOCopyBytesGracefully(c, resp, responseBody)
+	return &usage, nil
+}
+
+func GeminiResponsesStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	responseID := helper.GetResponseID(c)
+	created := common.GetTimestamp()
+	state := relayconvert.NewChatToResponsesStreamState(responseID, info.UpstreamModelName)
+	state.Created = created
+	finishReason := constant.FinishReasonStop
+	toolCallIndexByChoice := make(map[int]map[string]int)
+	nextToolCallIndexByChoice := make(map[int]int)
+	var streamErr *types.NewAPIError
+
+	sendEvent := func(event relayconvert.ChatToResponsesStreamEvent) bool {
+		data, err := common.Marshal(event.Payload)
+		if err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
+			return false
+		}
+		helper.ResponseChunkData(c, dto.ResponsesStreamResponse{Type: event.Type}, string(data))
+		return true
+	}
+	sendChunk := func(chunk *dto.ChatCompletionsStreamResponse) bool {
+		events, err := relayconvert.ChatCompletionsStreamChunkToResponsesEvents(chunk, state)
+		if err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+			return false
+		}
+		for _, event := range events {
+			if !sendEvent(event) {
+				return false
+			}
+		}
+		return true
+	}
+
+	usage, err := geminiStreamHandler(c, info, resp, func(data string, geminiResponse *dto.GeminiChatResponse) bool {
+		response, isStop := streamResponseGeminiChat2OpenAI(geminiResponse)
+		response.Id = responseID
+		response.Created = created
+		response.Model = info.UpstreamModelName
+
+		if response.IsToolCall() {
+			finishReason = constant.FinishReasonToolCalls
+		}
+		for choiceIdx := range response.Choices {
+			choiceKey := response.Choices[choiceIdx].Index
+			for toolIdx := range response.Choices[choiceIdx].Delta.ToolCalls {
+				tool := &response.Choices[choiceIdx].Delta.ToolCalls[toolIdx]
+				if tool.ID == "" {
+					continue
+				}
+				indexByID := toolCallIndexByChoice[choiceKey]
+				if indexByID == nil {
+					indexByID = make(map[string]int)
+					toolCallIndexByChoice[choiceKey] = indexByID
+				}
+				if idx, ok := indexByID[tool.ID]; ok {
+					tool.SetIndex(idx)
+					continue
+				}
+				idx := nextToolCallIndexByChoice[choiceKey]
+				nextToolCallIndexByChoice[choiceKey] = idx + 1
+				indexByID[tool.ID] = idx
+				tool.SetIndex(idx)
+			}
+		}
+
+		if !sendChunk(response) {
+			return false
+		}
+		if isStop {
+			return sendChunk(helper.GenerateStopResponse(responseID, created, info.UpstreamModelName, finishReason))
+		}
+		return true
+	})
+	if err != nil {
+		return usage, err
+	}
+	if streamErr != nil {
+		return nil, streamErr
+	}
+
+	if usage != nil {
+		state.Usage = relayconvert.UsageFromChatUsage(usage)
+	}
+	for _, event := range relayconvert.FinalizeChatCompletionsStreamToResponses(state) {
+		if !sendEvent(event) {
+			return nil, streamErr
+		}
+	}
+	return usage, nil
+}

relay/channel/gemini/relay_responses_test.go

+package gemini
+
+import (
+	"bytes"
+	"errors"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/constant"
+	"github.com/QuantumNous/new-api/dto"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	relayconstant "github.com/QuantumNous/new-api/relay/constant"
+	"github.com/QuantumNous/new-api/types"
+	"github.com/gin-gonic/gin"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestGeminiResponsesHandlerReturnsOpenAIResponsesJSON(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/responses", nil)
+	c.Set(common.RequestIdKey, "gemini-responses-test")
+
+	info := newGeminiResponsesRelayInfo(false)
+	payload := dto.GeminiChatResponse{
+		Candidates: []dto.GeminiChatCandidate{
+			{
+				Content: dto.GeminiChatContent{
+					Role: "model",
+					Parts: []dto.GeminiPart{
+						{Text: "hello"},
+					},
+				},
+			},
+		},
+		UsageMetadata: dto.GeminiUsageMetadata{
+			PromptTokenCount:     2,
+			CandidatesTokenCount: 3,
+			TotalTokenCount:      5,
+		},
+	}
+	body, err := common.Marshal(payload)
+	require.NoError(t, err)
+
+	usage, newAPIError := GeminiResponsesHandler(c, info, &http.Response{
+		Body: io.NopCloser(bytes.NewReader(body)),
+	})
+	require.Nil(t, newAPIError)
+	require.NotNil(t, usage)
+	assert.Equal(t, 2, usage.PromptTokens)
+	assert.Equal(t, 3, usage.CompletionTokens)
+
+	got := recorder.Body.String()
+	assert.Contains(t, got, `"object":"response"`)
+	assert.Contains(t, got, `"status":"completed"`)
+	assert.Contains(t, got, `"type":"output_text"`)
+	assert.Contains(t, got, `"text":"hello"`)
+	assert.Contains(t, got, `"input_tokens":2`)
+	assert.Contains(t, got, `"output_tokens":3`)
+	assert.NotContains(t, got, `"choices"`)
+	assert.NotContains(t, got, `"candidates"`)
+}
+
+func TestGeminiResponsesHandlerClosesBodyOnReadError(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/responses", nil)
+	c.Set(common.RequestIdKey, "gemini-responses-read-error-test")
+
+	body := &failingReadCloser{}
+	usage, newAPIError := GeminiResponsesHandler(c, newGeminiResponsesRelayInfo(false), &http.Response{Body: body})
+
+	require.Nil(t, usage)
+	require.NotNil(t, newAPIError)
+	assert.True(t, body.closed)
+}
+
+func TestGeminiResponsesStreamHandlerReturnsOpenAIResponsesSSE(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/responses", nil)
+	c.Set(common.RequestIdKey, "gemini-responses-stream-test")
+
+	oldStreamingTimeout := constant.StreamingTimeout
+	constant.StreamingTimeout = 300
+	t.Cleanup(func() { constant.StreamingTimeout = oldStreamingTimeout })
+
+	info := newGeminiResponsesRelayInfo(true)
+	first := dto.GeminiChatResponse{
+		Candidates: []dto.GeminiChatCandidate{
+			{
+				Content: dto.GeminiChatContent{
+					Role: "model",
+					Parts: []dto.GeminiPart{
+						{Text: "hello"},
+					},
+				},
+			},
+		},
+		UsageMetadata: dto.GeminiUsageMetadata{
+			PromptTokenCount:     2,
+			CandidatesTokenCount: 3,
+			TotalTokenCount:      5,
+		},
+	}
+	stop := "STOP"
+	final := dto.GeminiChatResponse{
+		Candidates: []dto.GeminiChatCandidate{
+			{
+				FinishReason: &stop,
+				Content: dto.GeminiChatContent{
+					Role:  "model",
+					Parts: []dto.GeminiPart{{Text: ""}},
+				},
+			},
+		},
+		UsageMetadata: dto.GeminiUsageMetadata{
+			PromptTokenCount:     2,
+			CandidatesTokenCount: 3,
+			TotalTokenCount:      5,
+		},
+	}
+	firstData, err := common.Marshal(first)
+	require.NoError(t, err)
+	finalData, err := common.Marshal(final)
+	require.NoError(t, err)
+	streamBody := strings.Join([]string{
+		"data: " + string(firstData),
+		"",
+		"data: " + string(finalData),
+		"",
+		"data: [DONE]",
+		"",
+	}, "\n")
+
+	usage, newAPIError := GeminiResponsesStreamHandler(c, info, &http.Response{
+		Body: io.NopCloser(strings.NewReader(streamBody)),
+	})
+	require.Nil(t, newAPIError)
+	require.NotNil(t, usage)
+	assert.Equal(t, 5, usage.TotalTokens)
+
+	got := recorder.Body.String()
+	assert.Equal(t, "text/event-stream", recorder.Header().Get("Content-Type"))
+	assert.Contains(t, got, `event: response.created`)
+	assert.Contains(t, got, `event: response.output_text.delta`)
+	assert.Contains(t, got, `"delta":"hello"`)
+	assert.Contains(t, got, `event: response.completed`)
+	assert.Contains(t, got, `"input_tokens":2`)
+	assert.Contains(t, got, `"output_tokens":3`)
+	assert.NotContains(t, got, `"choices"`)
+	assert.NotContains(t, got, `"candidates"`)
+	requireOrderedGeminiResponsesSubstrings(t, got,
+		`event: response.created`,
+		`event: response.output_item.added`,
+		`event: response.output_text.delta`,
+		`event: response.output_text.done`,
+		`event: response.completed`,
+	)
+}
+
+func newGeminiResponsesRelayInfo(isStream bool) *relaycommon.RelayInfo {
+	return &relaycommon.RelayInfo{
+		IsStream:        isStream,
+		RelayMode:       relayconstant.RelayModeResponses,
+		RelayFormat:     types.RelayFormatOpenAIResponses,
+		RequestURLPath:  "/v1/responses",
+		DisablePing:     true,
+		OriginModelName: "gemini-test",
+		ChannelMeta: &relaycommon.ChannelMeta{
+			UpstreamModelName: "gemini-test",
+		},
+	}
+}
+
+type failingReadCloser struct {
+	closed bool
+}
+
+func (r *failingReadCloser) Read([]byte) (int, error) {
+	return 0, errors.New("read failed")
+}
+
+func (r *failingReadCloser) Close() error {
+	r.closed = true
+	return nil
+}
+
+func requireOrderedGeminiResponsesSubstrings(t *testing.T, s string, parts ...string) {
+	t.Helper()
+	offset := 0
+	for _, part := range parts {
+		idx := strings.Index(s[offset:], part)
+		require.NotEqualf(t, -1, idx, "missing %q after byte offset %d", part, offset)
+		offset += idx + len(part)
+	}
+}

relay/channel/openai/chat_via_responses.go

@@ -14,7 +14,7 @@ import (
 	relaycommon "github.com/QuantumNous/new-api/relay/common"
 	"github.com/QuantumNous/new-api/relay/helper"
 	"github.com/QuantumNous/new-api/service"
-	"github.com/QuantumNous/new-api/service/openaicompat"
+	"github.com/QuantumNous/new-api/service/relayconvert"
 	"github.com/QuantumNous/new-api/types"
 
 	"github.com/gin-gonic/gin"
@@ -78,7 +78,7 @@ func OaiResponsesToChatBufferedStreamHandler(c *gin.Context, info *relaycommon.R
 	}
 	defer service.CloseResponseBodyGracefully(resp)
 
-	accumulator := openaicompat.NewResponsesBufferedAccumulator()
+	accumulator := relayconvert.NewResponsesBufferedAccumulator()
 	var finalResponse *dto.OpenAIResponsesResponse
 	var streamErr *types.NewAPIError
 
@@ -184,7 +184,7 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 
 	responseId := helper.GetResponseID(c)
 	createAt := time.Now().Unix()
-	state := openaicompat.NewResponsesToChatStreamState(info.UpstreamModelName, false)
+	state := relayconvert.NewResponsesToChatStreamState(info.UpstreamModelName, false)
 	state.ID = responseId
 	state.Created = createAt
 	streamErr := (*types.NewAPIError)(nil)
@@ -243,7 +243,7 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 			return
 		}
 
-		chunks, err := openaicompat.ResponsesStreamEventToChatChunks(&streamResp, state)
+		chunks, err := relayconvert.ResponsesStreamEventToChatChunks(&streamResp, state)
 		if err != nil {
 			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
 			sr.Stop(streamErr)
@@ -270,7 +270,7 @@ func OaiResponsesToChatStreamHandler(c *gin.Context, info *relaycommon.RelayInfo
 	if info.RelayFormat == types.RelayFormatClaude && info.ClaudeConvertInfo != nil {
 		info.ClaudeConvertInfo.Usage = usage
 	}
-	for _, chunk := range openaicompat.FinalizeResponsesToChatStream(state) {
+	for _, chunk := range relayconvert.FinalizeResponsesToChatStream(state) {
 		if !sendChatChunk(chunk) {
 			return nil, streamErr
 		}

relay/channel/openai/chat_via_responses_test.go

@@ -116,6 +116,58 @@ func TestOaiResponsesToChatBufferedStreamHandlerReturnsJSONFromSSE(t *testing.T)
 	require.Contains(t, got, `"finish_reason":"tool_calls"`)
 }
 
+func TestOaiChatToResponsesStreamHandlerConvertsSSEOrderAndUsage(t *testing.T) {
+	oldMode := gin.Mode()
+	gin.SetMode(gin.TestMode)
+	t.Cleanup(func() { gin.SetMode(oldMode) })
+
+	oldTimeout := constant.StreamingTimeout
+	constant.StreamingTimeout = 30
+	t.Cleanup(func() { constant.StreamingTimeout = oldTimeout })
+
+	body := strings.Join([]string{
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}`,
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[{"index":0,"delta":{"content":"hello"},"finish_reason":null}]}`,
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[{"index":0,"delta":{"tool_calls":[{"index":0,"id":"call_1","type":"function","function":{"name":"lookup"}}]},"finish_reason":null}]}`,
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[{"index":0,"delta":{"tool_calls":[{"index":0,"function":{"arguments":"{\"q\":\"x\"}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[{"index":0,"delta":{},"finish_reason":"tool_calls"}]}`,
+		`data: {"id":"chatcmpl_1","object":"chat.completion.chunk","created":1710000000,"model":"gpt-test","choices":[],"usage":{"prompt_tokens":2,"completion_tokens":3,"total_tokens":5}}`,
+		`data: [DONE]`,
+		``,
+	}, "\n")
+
+	c, recorder, resp, info := newResponsesChatTestContext(t, body, true)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/responses", nil)
+
+	usage, err := OaiChatToResponsesStreamHandler(c, info, resp)
+	require.Nil(t, err)
+	require.NotNil(t, usage)
+	require.Equal(t, 2, usage.PromptTokens)
+	require.Equal(t, 3, usage.CompletionTokens)
+	require.Equal(t, 5, usage.TotalTokens)
+
+	got := recorder.Body.String()
+	require.Equal(t, "text/event-stream", recorder.Header().Get("Content-Type"))
+	require.Contains(t, got, `event: response.created`)
+	require.Contains(t, got, `event: response.output_text.delta`)
+	require.Contains(t, got, `"delta":"hello"`)
+	require.Contains(t, got, `event: response.function_call_arguments.delta`)
+	require.Contains(t, got, `"delta":"{\"q\":\"x\"}"`)
+	require.Contains(t, got, `event: response.completed`)
+	require.Contains(t, got, `"input_tokens":2`)
+	require.Contains(t, got, `"output_tokens":3`)
+	requireOrderedSubstrings(t, got,
+		`event: response.created`,
+		`event: response.output_item.added`,
+		`event: response.output_text.delta`,
+		`event: response.output_item.added`,
+		`event: response.function_call_arguments.delta`,
+		`event: response.output_text.done`,
+		`event: response.function_call_arguments.done`,
+		`event: response.completed`,
+	)
+}
+
 func requireOrderedSubstrings(t *testing.T, s string, parts ...string) {
 	t.Helper()
 

relay/channel/openai/responses_via_chat.go

+package openai
+
+import (
+	"fmt"
+	"io"
+	"net/http"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/dto"
+	"github.com/QuantumNous/new-api/logger"
+	relaycommon "github.com/QuantumNous/new-api/relay/common"
+	"github.com/QuantumNous/new-api/relay/helper"
+	"github.com/QuantumNous/new-api/service"
+	"github.com/QuantumNous/new-api/service/relayconvert"
+	"github.com/QuantumNous/new-api/types"
+	"github.com/gin-gonic/gin"
+)
+
+func OaiChatToResponsesHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	if resp == nil || resp.Body == nil {
+		return nil, types.NewOpenAIError(fmt.Errorf("invalid response"), types.ErrorCodeBadResponse, http.StatusInternalServerError)
+	}
+	defer service.CloseResponseBodyGracefully(resp)
+
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeReadResponseBodyFailed, http.StatusInternalServerError)
+	}
+
+	var chatResp dto.OpenAITextResponse
+	if err := common.Unmarshal(body, &chatResp); err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponseBody, http.StatusInternalServerError)
+	}
+	if oaiError := chatResp.GetOpenAIError(); oaiError != nil && oaiError.Type != "" {
+		return nil, types.WithOpenAIError(*oaiError, resp.StatusCode)
+	}
+
+	responseID := helper.GetResponseID(c)
+	responsesResp, usage, err := service.ChatCompletionsResponseToResponsesResponse(&chatResp, responseID)
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeBadResponseBody, http.StatusInternalServerError)
+	}
+	if usage == nil || usage.TotalTokens == 0 {
+		text := service.ExtractOutputTextFromResponses(responsesResp)
+		usage = service.ResponseText2Usage(c, text, info.UpstreamModelName, info.GetEstimatePromptTokens())
+		responsesResp.Usage = relayconvert.UsageFromChatUsage(usage)
+	}
+
+	responseBody, err := common.Marshal(responsesResp)
+	if err != nil {
+		return nil, types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
+	}
+
+	service.IOCopyBytesGracefully(c, resp, responseBody)
+	return usage, nil
+}
+
+func OaiChatToResponsesStreamHandler(c *gin.Context, info *relaycommon.RelayInfo, resp *http.Response) (*dto.Usage, *types.NewAPIError) {
+	if resp == nil || resp.Body == nil {
+		return nil, types.NewOpenAIError(fmt.Errorf("invalid response"), types.ErrorCodeBadResponse, http.StatusInternalServerError)
+	}
+	defer service.CloseResponseBodyGracefully(resp)
+
+	responseID := helper.GetResponseID(c)
+	state := relayconvert.NewChatToResponsesStreamState(responseID, info.UpstreamModelName)
+	streamErr := (*types.NewAPIError)(nil)
+
+	sendEvent := func(event relayconvert.ChatToResponsesStreamEvent) bool {
+		data, err := common.Marshal(event.Payload)
+		if err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeJsonMarshalFailed, http.StatusInternalServerError)
+			return false
+		}
+		helper.ResponseChunkData(c, dto.ResponsesStreamResponse{Type: event.Type}, string(data))
+		return true
+	}
+
+	helper.StreamScannerHandler(c, resp, info, func(data string, sr *helper.StreamResult) {
+		if streamErr != nil {
+			sr.Stop(streamErr)
+			return
+		}
+
+		var errorResp dto.OpenAITextResponse
+		if err := common.UnmarshalJsonStr(data, &errorResp); err == nil {
+			if oaiError := errorResp.GetOpenAIError(); oaiError != nil && oaiError.Type != "" {
+				streamErr = types.WithOpenAIError(*oaiError, resp.StatusCode)
+				sr.Stop(streamErr)
+				return
+			}
+		}
+
+		var chunk dto.ChatCompletionsStreamResponse
+		if err := common.UnmarshalJsonStr(data, &chunk); err != nil {
+			logger.LogError(c, "failed to unmarshal chat stream response: "+err.Error())
+			sr.Error(err)
+			return
+		}
+
+		events, err := relayconvert.ChatCompletionsStreamChunkToResponsesEvents(&chunk, state)
+		if err != nil {
+			streamErr = types.NewOpenAIError(err, types.ErrorCodeBadResponse, http.StatusInternalServerError)
+			sr.Stop(streamErr)
+			return
+		}
+		for _, event := range events {
+			if !sendEvent(event) {
+				sr.Stop(streamErr)
+				return
+			}
+		}
+	})
+
+	if streamErr != nil {
+		return nil, streamErr
+	}
+
+	usage := state.Usage
+	if usage == nil || usage.TotalTokens == 0 {
+		usage = service.ResponseText2Usage(c, state.UsageText(), info.UpstreamModelName, info.GetEstimatePromptTokens())
+		state.Usage = relayconvert.UsageFromChatUsage(usage)
+	}
+
+	for _, event := range relayconvert.FinalizeChatCompletionsStreamToResponses(state) {
+		if !sendEvent(event) {
+			return nil, streamErr
+		}
+	}
+
+	return usage, nil
+}

relay/chat_completions_via_responses.go

@@ -146,7 +146,7 @@ func chatCompletionsViaResponses(c *gin.Context, info *relaycommon.RelayInfo, ad
 
 	httpResp = resp.(*http.Response)
 	clientStream := info.IsStream
-	upstreamStream := strings.HasPrefix(httpResp.Header.Get("Content-Type"), "text/event-stream")
+	upstreamStream := isResponsesEventStreamContentType(httpResp.Header.Get("Content-Type"))
 	info.IsStream = clientStream || upstreamStream
 	if httpResp.StatusCode != http.StatusOK {
 		newApiErr := service.RelayErrorHandler(c.Request.Context(), httpResp, false)
@@ -179,3 +179,7 @@ func chatCompletionsViaResponses(c *gin.Context, info *relaycommon.RelayInfo, ad
 	}
 	return usage, nil
 }
+
+func isResponsesEventStreamContentType(contentType string) bool {
+	return strings.Contains(strings.ToLower(contentType), "text/event-stream")
+}

relay/chat_completions_via_responses_test.go

+package relay
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestIsResponsesEventStreamContentType(t *testing.T) {
+	tests := []struct {
+		name        string
+		contentType string
+		want        bool
+	}{
+		{name: "plain", contentType: "text/event-stream", want: true},
+		{name: "mixed case with charset", contentType: "Text/Event-Stream; charset=utf-8", want: true},
+		{name: "json", contentType: "application/json", want: false},
+		{name: "empty", contentType: "", want: false},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			assert.Equal(t, tt.want, isResponsesEventStreamContentType(tt.contentType))
+		})
+	}
+}

service/convert.go

@@ -628,12 +628,14 @@ func ResponseOpenAI2Claude(openAIResponse *dto.OpenAITextResponse, info *relayco
 			claudeContent.Type = "tool_use"
 			claudeContent.Id = toolUse.ID
 			claudeContent.Name = toolUse.Function.Name
-			var mapParams map[string]interface{}
-			if err := common.Unmarshal([]byte(toolUse.Function.Arguments), &mapParams); err == nil {
-				claudeContent.Input = mapParams
-			} else {
-				claudeContent.Input = toolUse.Function.Arguments
+			mapParams := map[string]interface{}{}
+			if strings.TrimSpace(toolUse.Function.Arguments) != "" {
+				var parsed map[string]interface{}
+				if err := common.Unmarshal([]byte(toolUse.Function.Arguments), &parsed); err == nil && parsed != nil {
+					mapParams = parsed
+				}
 			}
+			claudeContent.Input = mapParams
 			contents = append(contents, claudeContent)
 		}
 	}

service/openai_chat_responses_compat.go

@@ -2,21 +2,29 @@ package service
 
 import (
 	"github.com/QuantumNous/new-api/dto"
-	"github.com/QuantumNous/new-api/service/openaicompat"
+	"github.com/QuantumNous/new-api/service/relayconvert"
 )
 
 func ChatCompletionsRequestToResponsesRequest(req *dto.GeneralOpenAIRequest) (*dto.OpenAIResponsesRequest, error) {
-	return openaicompat.ChatCompletionsRequestToResponsesRequest(req)
+	return relayconvert.ChatCompletionsRequestToResponsesRequest(req)
+}
+
+func ResponsesRequestToChatCompletionsRequest(req *dto.OpenAIResponsesRequest) (*dto.GeneralOpenAIRequest, error) {
+	return relayconvert.ResponsesRequestToChatCompletionsRequest(req)
+}
+
+func ChatCompletionsResponseToResponsesResponse(resp *dto.OpenAITextResponse, id string) (*dto.OpenAIResponsesResponse, *dto.Usage, error) {
+	return relayconvert.ChatCompletionsResponseToResponsesResponse(resp, id)
 }
 
 func ResponsesResponseToChatCompletionsResponse(resp *dto.OpenAIResponsesResponse, id string) (*dto.OpenAITextResponse, *dto.Usage, error) {
-	return openaicompat.ResponsesResponseToChatCompletionsResponse(resp, id)
+	return relayconvert.ResponsesResponseToChatCompletionsResponse(resp, id)
 }
 
 func ResponsesFinishReasonFromStatus(resp *dto.OpenAIResponsesResponse) (string, bool) {
-	return openaicompat.ResponsesFinishReasonFromStatus(resp)
+	return relayconvert.ResponsesFinishReasonFromStatus(resp)
 }
 
 func ExtractOutputTextFromResponses(resp *dto.OpenAIResponsesResponse) string {
-	return openaicompat.ExtractOutputTextFromResponses(resp)
+	return relayconvert.ExtractOutputTextFromResponses(resp)
 }

service/openai_chat_responses_mode.go

 package service
 
 import (
-	"github.com/QuantumNous/new-api/service/openaicompat"
+	"github.com/QuantumNous/new-api/service/relayconvert"
 	"github.com/QuantumNous/new-api/setting/model_setting"
 )
 
 func ShouldChatCompletionsUseResponsesPolicy(policy model_setting.ChatCompletionsToResponsesPolicy, channelID int, channelType int, model string) bool {
-	return openaicompat.ShouldChatCompletionsUseResponsesPolicy(policy, channelID, channelType, model)
+	return relayconvert.ShouldChatCompletionsUseResponsesPolicy(policy, channelID, channelType, model)
 }
 
 func ShouldChatCompletionsUseResponsesGlobal(channelID int, channelType int, model string) bool {
-	return openaicompat.ShouldChatCompletionsUseResponsesGlobal(channelID, channelType, model)
+	return relayconvert.ShouldChatCompletionsUseResponsesGlobal(channelID, channelType, model)
 }

service/openaicompat/chat_responses_compat_test.go → service/relayconvert/chat_responses_compat_test.go

-package openaicompat
+package relayconvert
 
 import (
 	"testing"
@@ -182,6 +182,70 @@ func TestResponsesStreamEventToChatChunksUsesOutputIndexForToolArguments(t *test
 	assert.Equal(t, 3, state.Usage.TotalTokens)
 }
 
+func TestResponsesStreamEventToChatChunksDoesNotDuplicatePendingArgsWithOutputIndexAndItemID(t *testing.T) {
+	state := newTestResponsesStreamState()
+	outputIndex := 1
+
+	var chunks []dto.ChatCompletionsStreamResponse
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{Type: responsesEventCreated})...)
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{
+		Type:        responsesEventFunctionArgsDelta,
+		OutputIndex: &outputIndex,
+		ItemID:      "fc_1",
+		Delta:       `{"q":"x"}`,
+	})...)
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{
+		Type:        responsesEventOutputItemAdded,
+		OutputIndex: &outputIndex,
+		ItemID:      "fc_1",
+		Item: &dto.ResponsesOutput{
+			Type:   responsesOutputTypeFunctionCall,
+			ID:     "fc_1",
+			CallId: "call_1",
+			Name:   "lookup",
+		},
+	})...)
+
+	require.Len(t, chunks, 2)
+	tool := chunks[1].Choices[0].Delta.ToolCalls[0]
+	assert.Equal(t, "call_1", tool.ID)
+	assert.Equal(t, "lookup", tool.Function.Name)
+	assert.Equal(t, `{"q":"x"}`, tool.Function.Arguments)
+	assert.Empty(t, state.pendingArgsByOutputIndex)
+	assert.Empty(t, state.pendingArgsByItemID)
+}
+
+func TestResponsesStreamEventToChatChunksDrainsItemOnlyPendingArgsWhenOutputIndexArrives(t *testing.T) {
+	state := newTestResponsesStreamState()
+	outputIndex := 1
+
+	var chunks []dto.ChatCompletionsStreamResponse
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{Type: responsesEventCreated})...)
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{
+		Type:   responsesEventFunctionArgsDelta,
+		ItemID: "fc_1",
+		Delta:  `{"q":"x"}`,
+	})...)
+	chunks = append(chunks, mustStreamChunks(t, state, &dto.ResponsesStreamResponse{
+		Type:        responsesEventOutputItemAdded,
+		OutputIndex: &outputIndex,
+		ItemID:      "fc_1",
+		Item: &dto.ResponsesOutput{
+			Type:   responsesOutputTypeFunctionCall,
+			CallId: "call_1",
+			Name:   "lookup",
+		},
+	})...)
+
+	require.Len(t, chunks, 2)
+	tool := chunks[1].Choices[0].Delta.ToolCalls[0]
+	assert.Equal(t, "call_1", tool.ID)
+	assert.Equal(t, "lookup", tool.Function.Name)
+	assert.Equal(t, `{"q":"x"}`, tool.Function.Arguments)
+	assert.Empty(t, state.pendingArgsByOutputIndex)
+	assert.Empty(t, state.pendingArgsByItemID)
+}
+
 func TestResponsesStreamEventToChatChunksCustomToolAndReasoning(t *testing.T) {
 	state := newTestResponsesStreamState()
 	outputIndex := 0
@@ -313,6 +377,165 @@ func TestResponsesBufferedAccumulatorSupplementsEmptyTerminalOutput(t *testing.T
 	assert.Equal(t, `{"q":"x"}`, toolCalls[0].Function.Arguments)
 }
 
+func TestResponsesBufferedAccumulatorDoesNotDuplicatePendingArgsWithOutputIndexAndItemID(t *testing.T) {
+	acc := NewResponsesBufferedAccumulator()
+	outputIndex := 1
+	acc.ProcessEvent(&dto.ResponsesStreamResponse{
+		Type:        responsesEventFunctionArgsDelta,
+		OutputIndex: &outputIndex,
+		ItemID:      "fc_1",
+		Delta:       `{"q":"x"}`,
+	})
+	acc.ProcessEvent(&dto.ResponsesStreamResponse{
+		Type:        responsesEventOutputItemAdded,
+		OutputIndex: &outputIndex,
+		ItemID:      "fc_1",
+		Item: &dto.ResponsesOutput{
+			Type:   responsesOutputTypeFunctionCall,
+			ID:     "fc_1",
+			CallId: "call_1",
+			Name:   "lookup",
+		},
+	})
+
+	resp := &dto.OpenAIResponsesResponse{
+		Status: []byte(`"completed"`),
+		Model:  "gpt-test",
+	}
+	acc.SupplementResponseOutput(resp)
+
+	chat, _, err := ResponsesResponseToChatCompletionsResponse(resp, "chatcmpl_1")
+	require.NoError(t, err)
+	toolCalls := chat.Choices[0].Message.ParseToolCalls()
+	require.Len(t, toolCalls, 1)
+	assert.Equal(t, `{"q":"x"}`, toolCalls[0].Function.Arguments)
+	assert.Empty(t, acc.pendingByOutputIndex)
+	assert.Empty(t, acc.pendingByItemID)
+}
+
+func TestChatCompletionsResponseToResponsesPreservesTextToolCallsAndUsage(t *testing.T) {
+	chat := &dto.OpenAITextResponse{
+		Id:      "chatcmpl_1",
+		Model:   "gpt-test",
+		Created: 456,
+		Choices: []dto.OpenAITextResponseChoice{
+			{
+				Message:      assistantMessageWithTool("I will call.", "call_1", "lookup", `{"q":"x"}`),
+				FinishReason: "tool_calls",
+			},
+		},
+		Usage: dto.Usage{PromptTokens: 3, CompletionTokens: 5, TotalTokens: 8},
+	}
+
+	resp, usage, err := ChatCompletionsResponseToResponsesResponse(chat, "resp_1")
+	require.NoError(t, err)
+	require.NotNil(t, usage)
+
+	assert.Equal(t, "resp_1", resp.ID)
+	assert.Equal(t, "response", resp.Object)
+	assert.Equal(t, `"completed"`, string(resp.Status))
+	assert.Equal(t, 3, resp.Usage.InputTokens)
+	assert.Equal(t, 5, resp.Usage.OutputTokens)
+	require.Len(t, resp.Output, 2)
+	assert.Equal(t, responsesOutputTypeMessage, resp.Output[0].Type)
+	assert.Equal(t, "I will call.", resp.Output[0].Content[0].Text)
+	assert.Equal(t, responsesOutputTypeFunctionCall, resp.Output[1].Type)
+	assert.Equal(t, "call_1", resp.Output[1].CallId)
+	assert.Equal(t, "lookup", resp.Output[1].Name)
+	assert.Equal(t, `"{\"q\":\"x\"}"`, string(resp.Output[1].Arguments))
+}
+
+func TestChatCompletionsResponseToResponsesMapsIncompleteFinishReasons(t *testing.T) {
+	tests := []struct {
+		name         string
+		finishReason string
+		wantReason   string
+	}{
+		{name: "length", finishReason: "length", wantReason: responsesIncompleteReasonMaxTokens},
+		{name: "content filter", finishReason: "content_filter", wantReason: responsesIncompleteReasonContentFilter},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			resp, _, err := ChatCompletionsResponseToResponsesResponse(&dto.OpenAITextResponse{
+				Id:    "chatcmpl_1",
+				Model: "gpt-test",
+				Choices: []dto.OpenAITextResponseChoice{
+					{
+						Message:      dto.Message{Role: "assistant", Content: "partial"},
+						FinishReason: tt.finishReason,
+					},
+				},
+			}, "resp_1")
+			require.NoError(t, err)
+
+			assert.Equal(t, `"incomplete"`, string(resp.Status))
+			require.NotNil(t, resp.IncompleteDetails)
+			assert.Equal(t, tt.wantReason, resp.IncompleteDetails.Reason)
+			require.Len(t, resp.Output, 1)
+			assert.Equal(t, "incomplete", resp.Output[0].Status)
+		})
+	}
+}
+
+func TestChatCompletionsStreamToResponsesEventsAggregatesUsageAndToolArgs(t *testing.T) {
+	state := NewChatToResponsesStreamState("resp_1", "gpt-test")
+	state.Created = 123
+	toolIndex := 0
+
+	var events []ChatToResponsesStreamEvent
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Id:      "chatcmpl_1",
+		Model:   "gpt-test",
+		Created: 123,
+		Choices: []dto.ChatCompletionsStreamResponseChoice{
+			{Index: 0, Delta: dto.ChatCompletionsStreamResponseChoiceDelta{Role: "assistant"}},
+		},
+	})...)
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Choices: []dto.ChatCompletionsStreamResponseChoice{
+			{Index: 0, Delta: dto.ChatCompletionsStreamResponseChoiceDelta{Content: lo.ToPtr("hello")}},
+		},
+	})...)
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Choices: []dto.ChatCompletionsStreamResponseChoice{
+			{Index: 0, Delta: dto.ChatCompletionsStreamResponseChoiceDelta{ToolCalls: []dto.ToolCallResponse{
+				{Index: &toolIndex, ID: "call_1", Type: "function", Function: dto.FunctionResponse{Name: "lookup"}},
+			}}},
+		},
+	})...)
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Choices: []dto.ChatCompletionsStreamResponseChoice{
+			{Index: 0, Delta: dto.ChatCompletionsStreamResponseChoiceDelta{ToolCalls: []dto.ToolCallResponse{
+				{Index: &toolIndex, Function: dto.FunctionResponse{Arguments: `{"q":"x"}`}},
+			}}},
+		},
+	})...)
+	finishReason := "tool_calls"
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Choices: []dto.ChatCompletionsStreamResponseChoice{
+			{Index: 0, FinishReason: &finishReason},
+		},
+	})...)
+	events = append(events, mustResponsesEventsFromChatChunk(t, state, &dto.ChatCompletionsStreamResponse{
+		Usage: &dto.Usage{PromptTokens: 2, CompletionTokens: 4, TotalTokens: 6},
+	})...)
+	events = append(events, FinalizeChatCompletionsStreamToResponses(state)...)
+
+	require.Len(t, events, 10)
+	assert.Equal(t, responsesEventCreated, events[0].Type)
+	assert.Equal(t, responsesEventOutputTextDelta, events[2].Type)
+	assert.Equal(t, "hello", events[2].Payload.Delta)
+	assert.Equal(t, responsesEventFunctionArgsDelta, events[4].Type)
+	assert.Equal(t, `{"q":"x"}`, events[4].Payload.Delta)
+	assert.Equal(t, responsesEventCompleted, events[9].Type)
+	require.NotNil(t, events[9].Payload.Response)
+	assert.Equal(t, 6, events[9].Payload.Response.Usage.TotalTokens)
+	require.Len(t, events[9].Payload.Response.Output, 2)
+	assert.Equal(t, "hello", events[9].Payload.Response.Output[0].Content[0].Text)
+	assert.Equal(t, `"{\"q\":\"x\"}"`, string(events[9].Payload.Response.Output[1].Arguments))
+}
+
 func assistantMessageWithTool(content string, id string, name string, args string) dto.Message {
 	msg := dto.Message{Role: "assistant", Content: content}
 	msg.SetToolCalls([]dto.ToolCallRequest{
@@ -341,3 +564,10 @@ func mustStreamChunks(t *testing.T, state *ResponsesToChatStreamState, event *dt
 	require.NoError(t, err)
 	return chunks
 }
+
+func mustResponsesEventsFromChatChunk(t *testing.T, state *ChatToResponsesStreamState, chunk *dto.ChatCompletionsStreamResponse) []ChatToResponsesStreamEvent {
+	t.Helper()
+	events, err := ChatCompletionsStreamChunkToResponsesEvents(chunk, state)
+	require.NoError(t, err)
+	return events
+}

service/openaicompat/chat_to_responses.go → service/relayconvert/chat_to_responses.go

-package openaicompat
+package relayconvert
 
 import (
 	"encoding/json"

service/openaicompat/policy.go → service/relayconvert/policy.go

-package openaicompat
+package relayconvert
 
 import "github.com/QuantumNous/new-api/setting/model_setting"
 

service/openaicompat/regex.go → service/relayconvert/regex.go

-package openaicompat
+package relayconvert
 
 import (
 	"regexp"

service/relayconvert/responses_request_to_chat_test.go

+package relayconvert
+
+import (
+	"testing"
+
+	"github.com/QuantumNous/new-api/common"
+	"github.com/QuantumNous/new-api/dto"
+	"github.com/samber/lo"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"github.com/tidwall/gjson"
+)
+
+func TestResponsesRequestToChatCompletionsRequestInstructionsAndScalarInput(t *testing.T) {
+	stream := true
+	temperature := 0.0
+	topP := 0.9
+	maxOutputTokens := uint(128)
+	parallelToolCalls := true
+
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model:                "gpt-test",
+		Instructions:         mustRawMessage(t, "system rules"),
+		Input:                mustRawMessage(t, "hello"),
+		Stream:               &stream,
+		StreamOptions:        &dto.StreamOptions{IncludeUsage: true},
+		MaxOutputTokens:      &maxOutputTokens,
+		Temperature:          &temperature,
+		TopP:                 &topP,
+		User:                 mustRawMessage(t, "user-1"),
+		Store:                mustRawMessage(t, false),
+		Metadata:             mustRawMessage(t, map[string]any{"trace": "abc"}),
+		ParallelToolCalls:    mustRawMessage(t, parallelToolCalls),
+		PromptCacheKey:       mustRawMessage(t, "cache-key"),
+		PromptCacheRetention: mustRawMessage(t, "24h"),
+		Reasoning:            &dto.Reasoning{Effort: "medium"},
+	})
+	require.NoError(t, err)
+
+	assert.Equal(t, "gpt-test", got.Model)
+	require.Len(t, got.Messages, 2)
+	assert.Equal(t, dto.Message{Role: "system", Content: "system rules"}, got.Messages[0])
+	assert.Equal(t, dto.Message{Role: "user", Content: "hello"}, got.Messages[1])
+	assert.Same(t, &stream, got.Stream)
+	require.NotNil(t, got.StreamOptions)
+	assert.True(t, got.StreamOptions.IncludeUsage)
+	assert.Equal(t, maxOutputTokens, lo.FromPtr(got.MaxCompletionTokens))
+	assert.Equal(t, 0.0, lo.FromPtr(got.Temperature))
+	assert.Equal(t, 0.9, lo.FromPtr(got.TopP))
+	assert.True(t, lo.FromPtr(got.ParallelTooCalls))
+	assert.Equal(t, "cache-key", got.PromptCacheKey)
+	assert.Equal(t, "medium", got.ReasoningEffort)
+	assert.Equal(t, `"user-1"`, string(got.User))
+	assert.Equal(t, `false`, string(got.Store))
+	assert.Equal(t, "abc", gjson.GetBytes(got.Metadata, "trace").String())
+}
+
+func TestResponsesRequestToChatCompletionsRequestMultimodalInput(t *testing.T) {
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model: "gpt-test",
+		Input: mustRawMessage(t, []map[string]any{
+			{
+				"role": "user",
+				"content": []map[string]any{
+					{"type": "input_text", "text": "look"},
+					{"type": "input_image", "image_url": "https://example.test/a.png", "detail": "low"},
+					{"type": "input_file", "file_id": "file_1", "filename": "a.txt"},
+					{"type": "input_audio", "input_audio": map[string]any{"data": "abc", "format": "wav"}},
+					{"type": "input_video", "video_url": map[string]any{"url": "https://example.test/v.mp4"}},
+				},
+			},
+		}),
+	})
+	require.NoError(t, err)
+
+	require.Len(t, got.Messages, 1)
+	assert.Equal(t, "user", got.Messages[0].Role)
+	parts := got.Messages[0].ParseContent()
+	require.Len(t, parts, 5)
+	assert.Equal(t, dto.ContentTypeText, parts[0].Type)
+	assert.Equal(t, "look", parts[0].Text)
+	assert.Equal(t, dto.ContentTypeImageURL, parts[1].Type)
+	assert.Equal(t, "https://example.test/a.png", parts[1].GetImageMedia().Url)
+	assert.Equal(t, dto.ContentTypeFile, parts[2].Type)
+	assert.Equal(t, "file_1", parts[2].GetFile().FileId)
+	assert.Equal(t, dto.ContentTypeInputAudio, parts[3].Type)
+	assert.Equal(t, "wav", parts[3].GetInputAudio().Format)
+	assert.Equal(t, dto.ContentTypeVideoUrl, parts[4].Type)
+	assert.Equal(t, "https://example.test/v.mp4", parts[4].GetVideoUrl().Url)
+}
+
+func TestResponsesRequestToChatCompletionsRequestAssistantTextAndFunctionCallCoexist(t *testing.T) {
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model: "gpt-test",
+		Input: mustRawMessage(t, []map[string]any{
+			{
+				"role": "assistant",
+				"content": []map[string]any{
+					{"type": "output_text", "text": "I will call."},
+				},
+			},
+			{
+				"type":      "function_call",
+				"call_id":   "call_1",
+				"name":      "lookup",
+				"arguments": map[string]any{"q": "x"},
+			},
+			{
+				"type":    "function_call_output",
+				"call_id": "call_1",
+				"output":  map[string]any{"ok": true},
+			},
+		}),
+	})
+	require.NoError(t, err)
+
+	require.Len(t, got.Messages, 2)
+	assert.Equal(t, "assistant", got.Messages[0].Role)
+	assert.Equal(t, "I will call.", got.Messages[0].StringContent())
+	toolCalls := got.Messages[0].ParseToolCalls()
+	require.Len(t, toolCalls, 1)
+	assert.Equal(t, "call_1", toolCalls[0].ID)
+	assert.Equal(t, "function", toolCalls[0].Type)
+	assert.Equal(t, "lookup", toolCalls[0].Function.Name)
+	assert.JSONEq(t, `{"q":"x"}`, toolCalls[0].Function.Arguments)
+	assert.Equal(t, "tool", got.Messages[1].Role)
+	assert.Equal(t, "call_1", got.Messages[1].ToolCallId)
+	assert.JSONEq(t, `{"ok":true}`, got.Messages[1].StringContent())
+}
+
+func TestResponsesRequestToChatCompletionsRequestOnlyFunctionCallCreatesAssistant(t *testing.T) {
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model: "gpt-test",
+		Input: mustRawMessage(t, []map[string]any{
+			{
+				"type":      "function_call",
+				"call_id":   "call_1",
+				"name":      "lookup",
+				"arguments": `{"q":"x"}`,
+			},
+		}),
+	})
+	require.NoError(t, err)
+
+	require.Len(t, got.Messages, 1)
+	assert.Equal(t, "assistant", got.Messages[0].Role)
+	assert.Nil(t, got.Messages[0].Content)
+	toolCalls := got.Messages[0].ParseToolCalls()
+	require.Len(t, toolCalls, 1)
+	assert.Equal(t, `{"q":"x"}`, toolCalls[0].Function.Arguments)
+}
+
+func TestResponsesRequestToChatCompletionsRequestToolsToolChoiceAndTextFormat(t *testing.T) {
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model: "gpt-test",
+		Input: mustRawMessage(t, "hello"),
+		Tools: mustRawMessage(t, []map[string]any{
+			{
+				"type":        "function",
+				"name":        "lookup",
+				"description": "Lookup data",
+				"parameters": map[string]any{
+					"type": "object",
+					"properties": map[string]any{
+						"q": map[string]any{"type": "string"},
+					},
+				},
+			},
+		}),
+		ToolChoice: mustRawMessage(t, map[string]any{
+			"type": "function",
+			"name": "lookup",
+		}),
+		Text: mustRawMessage(t, map[string]any{
+			"format": map[string]any{
+				"type":   "json_schema",
+				"name":   "answer",
+				"schema": map[string]any{"type": "object"},
+				"strict": true,
+			},
+		}),
+	})
+	require.NoError(t, err)
+
+	require.Len(t, got.Tools, 1)
+	assert.Equal(t, "function", got.Tools[0].Type)
+	assert.Equal(t, "lookup", got.Tools[0].Function.Name)
+	assert.Equal(t, "Lookup data", got.Tools[0].Function.Description)
+	assert.Equal(t, "object", got.Tools[0].Function.Parameters.(map[string]any)["type"])
+	assert.Equal(t, map[string]any{
+		"type": "function",
+		"function": map[string]any{
+			"name": "lookup",
+		},
+	}, got.ToolChoice)
+	require.NotNil(t, got.ResponseFormat)
+	assert.Equal(t, "json_schema", got.ResponseFormat.Type)
+	assert.Equal(t, "answer", gjson.GetBytes(got.ResponseFormat.JsonSchema, "name").String())
+	assert.True(t, gjson.GetBytes(got.ResponseFormat.JsonSchema, "strict").Bool())
+}
+
+func TestResponsesRequestToChatCompletionsRequestCustomToolCallPreservesRawShape(t *testing.T) {
+	got, err := ResponsesRequestToChatCompletionsRequest(&dto.OpenAIResponsesRequest{
+		Model: "gpt-test",
+		Input: mustRawMessage(t, []map[string]any{
+			{
+				"type":    "custom_tool_call",
+				"call_id": "call_custom",
+				"name":    "apply_patch",
+				"input":   "patch body",
+			},
+		}),
+	})
+	require.NoError(t, err)
+
+	require.Len(t, got.Messages, 1)
+	toolCalls := got.Messages[0].ParseToolCalls()
+	require.Len(t, toolCalls, 1)
+	assert.Equal(t, dto.CustomType, toolCalls[0].Type)
+	assert.Equal(t, "call_custom", toolCalls[0].ID)
+	assert.Equal(t, "apply_patch", toolCalls[0].Function.Name)
+	assert.Equal(t, "patch body", toolCalls[0].Function.Arguments)
+	assert.Equal(t, "custom_tool_call", gjson.GetBytes(toolCalls[0].Custom, "type").String())
+	assert.Equal(t, "patch body", gjson.GetBytes(toolCalls[0].Custom, "input").String())
+}
+
+func TestResponsesRequestToChatCompletionsRequestRejectsStatefulFields(t *testing.T) {
+	tests := []struct {
+		name string
+		req  *dto.OpenAIResponsesRequest
+		want string
+	}{
+		{
+			name: "conversation",
+			req:  &dto.OpenAIResponsesRequest{Model: "gpt-test", Conversation: mustRawMessage(t, "conv_1")},
+			want: "conversation",
+		},
+		{
+			name: "previous response",
+			req:  &dto.OpenAIResponsesRequest{Model: "gpt-test", PreviousResponseID: "resp_1"},
+			want: "previous_response_id",
+		},
+		{
+			name: "prompt",
+			req:  &dto.OpenAIResponsesRequest{Model: "gpt-test", Prompt: mustRawMessage(t, map[string]any{"id": "pmpt_1"})},
+			want: "prompt",
+		},
+		{
+			name: "context management",
+			req:  &dto.OpenAIResponsesRequest{Model: "gpt-test", ContextManagement: mustRawMessage(t, map[string]any{"type": "auto"})},
+			want: "context_management",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			_, err := ResponsesRequestToChatCompletionsRequest(tt.req)
+			require.Error(t, err)
+			assert.Contains(t, err.Error(), tt.want)
+			assert.Contains(t, err.Error(), "stateful fields")
+		})
+	}
+}
+
+func mustRawMessage(t *testing.T, value any) []byte {
+	t.Helper()
+	raw, err := common.Marshal(value)
+	require.NoError(t, err)
+	return raw
+}

service/openaicompat/responses_to_chat.go → service/relayconvert/responses_to_chat.go

-package openaicompat
+package relayconvert
 
 import (
 	"errors"
@@ -438,8 +438,7 @@ func (s *ResponsesToChatStreamState) toolArgumentsDelta(event *dto.ResponsesStre
 	if tool == nil {
 		if event.OutputIndex != nil {
 			s.pendingArgsByOutputIndex[*event.OutputIndex] += event.Delta
-		}
-		if itemID := strings.TrimSpace(event.ItemID); itemID != "" {
+		} else if itemID := strings.TrimSpace(event.ItemID); itemID != "" {
 			s.pendingArgsByItemID[itemID] += event.Delta
 		}
 		return nil
@@ -485,7 +484,7 @@ func (s *ResponsesToChatStreamState) ensureToolForEvent(event *dto.ResponsesStre
 			delete(s.pendingArgsByOutputIndex, *event.OutputIndex)
 		}
 	}
-	if itemID := strings.TrimSpace(event.Item.ID); itemID != "" {
+	if itemID := responseStreamEventItemID(event); itemID != "" {
 		tool.ItemID = itemID
 		s.itemIDToKey[itemID] = key
 		if pending := s.pendingArgsByItemID[itemID]; pending != "" {
@@ -558,7 +557,7 @@ func (s *ResponsesToChatStreamState) ensureFallbackToolForEvent(event *dto.Respo
 			delete(s.pendingArgsByOutputIndex, *event.OutputIndex)
 		}
 	}
-	if itemID := strings.TrimSpace(event.ItemID); itemID != "" {
+	if itemID := responseStreamEventItemID(event); itemID != "" {
 		tool.ItemID = itemID
 		s.itemIDToKey[itemID] = key
 		if pending := s.pendingArgsByItemID[itemID]; pending != "" {
@@ -789,8 +788,7 @@ func (a *ResponsesBufferedAccumulator) ProcessEvent(event *dto.ResponsesStreamRe
 		}
 		if event.OutputIndex != nil {
 			a.pendingByOutputIndex[*event.OutputIndex] += event.Delta
-		}
-		if itemID := strings.TrimSpace(event.ItemID); itemID != "" {
+		} else if itemID := strings.TrimSpace(event.ItemID); itemID != "" {
 			a.pendingByItemID[itemID] += event.Delta
 		}
 	}
@@ -928,6 +926,18 @@ func isResponsesToolOutputType(outputType string) bool {
 	return outputType == responsesOutputTypeFunctionCall || outputType == responsesOutputTypeCustomToolCall
 }
 
+func responseStreamEventItemID(event *dto.ResponsesStreamResponse) string {
+	if event == nil {
+		return ""
+	}
+	if event.Item != nil {
+		if itemID := strings.TrimSpace(event.Item.ID); itemID != "" {
+			return itemID
+		}
+	}
+	return strings.TrimSpace(event.ItemID)
+}
+
 func fallbackToolKey(itemID string, callID string, outputIndex *int) string {
 	if outputIndex != nil {
 		return fmt.Sprintf("output:%d", *outputIndex)

web/default/src/features/channels/lib/advanced-custom.ts

@@ -44,6 +44,10 @@ export const ADVANCED_CUSTOM_CONVERTER_OPTIONS: Array<{
     label: 'OpenAI Chat to OpenAI Responses',
   },
   {
+    value: 'openai_responses_to_openai_chat_completions',
+    label: 'OpenAI Responses to OpenAI Chat',
+  },
+  {
     value: 'gemini_generate_content_to_openai_chat_completions',
     label: 'Gemini Generate Content to OpenAI Chat',
   },
@@ -326,6 +330,9 @@ export function getAdvancedCustomUpstreamPathPlaceholder(
   if (converter === 'openai_chat_completions_to_anthropic_messages') {
     return '/v1/messages'
   }
+  if (converter === 'openai_responses_to_openai_chat_completions') {
+    return '/v1/chat/completions'
+  }
   return '/v1/chat/completions'
 }
 
@@ -573,6 +580,9 @@ function isConverterPathAllowed(
   ) {
     return incomingPath === '/v1/chat/completions'
   }
+  if (converter === 'openai_responses_to_openai_chat_completions') {
+    return incomingPath === '/v1/responses'
+  }
   return (
     incomingPath.includes(':generateContent') ||
     incomingPath.includes(':streamGenerateContent')

web/default/src/features/channels/types.ts

@@ -131,6 +131,7 @@ export type AdvancedCustomConverter =
   | 'anthropic_messages_to_openai_chat_completions'
   | 'openai_chat_completions_to_anthropic_messages'
   | 'openai_chat_completions_to_openai_responses'
+  | 'openai_responses_to_openai_chat_completions'
   | 'gemini_generate_content_to_openai_chat_completions'
   | 'openai_chat_completions_to_gemini_generate_content'
 

web/default/src/i18n/locales/en.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "OpenAI Rerank",
     "OpenAI Responses": "OpenAI Responses",
     "OpenAI Responses Compact": "OpenAI Responses Compact",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses to OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI, Anthropic, etc.",
     "OpenAI, Anthropic, Google, etc.": "OpenAI, Anthropic, Google, etc.",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "Resetting...",
     "Resolve Conflicts": "Resolve Conflicts",
     "Resource Configuration": "Resource Configuration",
-    "Responding...": "Responding...",
     "Resources": "Resources",
+    "Responding...": "Responding...",
     "Response": "Response",
     "Response Time": "Response Time",
     "Response time: {{duration}}": "Response time: {{duration}}",

web/default/src/i18n/locales/fr.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "OpenAI Rerank",
     "OpenAI Responses": "OpenAI Responses",
     "OpenAI Responses Compact": "OpenAI Responses Compact",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses vers OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI, Anthropic, etc.",
     "OpenAI, Anthropic, Google, etc.": "OpenAI, Anthropic, Google, etc.",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "Réinitialisation...",
     "Resolve Conflicts": "Résoudre les conflits",
     "Resource Configuration": "Configuration des ressources",
-    "Responding...": "Réponse en cours...",
     "Resources": "Ressources",
+    "Responding...": "Réponse en cours...",
     "Response": "Réponse",
     "Response Time": "Temps de réponse",
     "Response time: {{duration}}": "Temps de réponse : {{duration}}",

web/default/src/i18n/locales/ja.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "OpenAI 再ランク付け",
     "OpenAI Responses": "OpenAI レスポンス",
     "OpenAI Responses Compact": "OpenAI レスポンス圧縮",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses から OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI、Anthropicなど",
     "OpenAI, Anthropic, Google, etc.": "OpenAI、Anthropic、Googleなど",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "リセット中...",
     "Resolve Conflicts": "競合を解決",
     "Resource Configuration": "リソース設定",
-    "Responding...": "応答中...",
     "Resources": "リソース",
+    "Responding...": "応答中...",
     "Response": "レスポンス",
     "Response Time": "応答時間",
     "Response time: {{duration}}": "応答時間: {{duration}}",

web/default/src/i18n/locales/ru.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "Реранжирование OpenAI",
     "OpenAI Responses": "Ответы OpenAI",
     "OpenAI Responses Compact": "Компактные ответы OpenAI",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses в OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI, Anthropic и т.д.",
     "OpenAI, Anthropic, Google, etc.": "OpenAI, Anthropic, Google и т.д.",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "Сброс...",
     "Resolve Conflicts": "Разрешить конфликты",
     "Resource Configuration": "Конфигурация ресурсов",
-    "Responding...": "Отвечаем...",
     "Resources": "Ресурсы",
+    "Responding...": "Отвечаем...",
     "Response": "Ответ",
     "Response Time": "Время ответа",
     "Response time: {{duration}}": "Время ответа: {{duration}}",

web/default/src/i18n/locales/vi.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "OpenAI Rerank",
     "OpenAI Responses": "OpenAI Responses",
     "OpenAI Responses Compact": "OpenAI Responses Compact",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses sang OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI, Anthropic, v.v.",
     "OpenAI, Anthropic, Google, etc.": "OpenAI, Anthropic, Google, v.v.",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "Đang đặt lại...",
     "Resolve Conflicts": "Giải quyết Xung đột",
     "Resource Configuration": "Cấu hình tài nguyên",
-    "Responding...": "Đang phản hồi...",
     "Resources": "Tài nguyên",
+    "Responding...": "Đang phản hồi...",
     "Response": "Phản hồi",
     "Response Time": "Thời gian phản hồi",
     "Response time: {{duration}}": "Thời gian phản hồi: {{duration}}",

web/default/src/i18n/locales/zh.json

@@ -2995,6 +2995,7 @@
     "OpenAI Rerank": "OpenAI 重排序",
     "OpenAI Responses": "OpenAI 响应",
     "OpenAI Responses Compact": "OpenAI 响应压缩",
+    "OpenAI Responses to OpenAI Chat": "OpenAI Responses 到 OpenAI Chat",
     "OpenAI, Anthropic, etc.": "OpenAI、Anthropic 等",
     "OpenAI, Anthropic, Google, etc.": "OpenAI、Anthropic、Google 等",
     "OpenAIMax": "OpenAIMax",
@@ -3675,8 +3676,8 @@
     "Resetting...": "重置中...",
     "Resolve Conflicts": "解决冲突",
     "Resource Configuration": "资源配置",
-    "Responding...": "正在回复...",
     "Resources": "资源",
+    "Responding...": "正在回复...",
     "Response": "响应",
     "Response Time": "响应时间",
     "Response time: {{duration}}": "响应时间：{{duration}}",