feat: user provided api key & bug fix.

.github/workflows/release.yml

@@ -22,6 +22,7 @@ jobs:
         run: |
           export PD_API_ENDPOINT=https://app.paperdebugger.com
           export BETA_BUILD=false
+          export GRAFANA_API_KEY=${{ secrets.GRAFANA_API_KEY }}
           cd webapp/_webapp
           npm install
           npm run build

go.mod

@@ -13,7 +13,7 @@ require (
 	github.com/google/wire v0.7.0
 	github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.2
 	github.com/joho/godotenv v1.5.1
-	github.com/openai/openai-go/v2 v2.1.1
+	github.com/openai/openai-go/v2 v2.7.1
 	github.com/samber/lo v1.51.0
 	github.com/stretchr/testify v1.10.0
 	go.mongodb.org/mongo-driver/v2 v2.3.0

go.sum

@@ -88,8 +88,8 @@ github.com/modern-go/reflect2 v1.0.2 h1:xBagoLtFs94CBntxluKeaWgTMpvLxC4ur3nMaC9G
 github.com/modern-go/reflect2 v1.0.2/go.mod h1:yWuevngMOJpCy52FWWMvUC8ws7m/LJsjYzDa0/r8luk=
 github.com/muesli/termenv v0.16.0 h1:S5AlUN9dENB57rsbnkPyfdGuWIlkmzJjbFf0Tf5FWUc=
 github.com/muesli/termenv v0.16.0/go.mod h1:ZRfOIKPFDYQoDFF4Olj7/QJbW60Ol/kL1pU3VfY/Cnk=
-github.com/openai/openai-go/v2 v2.1.1 h1:/RMA/V3D+yF/Cc4jHXFt6lkqSOWRf5roRi+DvZaDYQI=
-github.com/openai/openai-go/v2 v2.1.1/go.mod h1:sIUkR+Cu/PMUVkSKhkk742PRURkQOCFhiwJ7eRSBqmk=
+github.com/openai/openai-go/v2 v2.7.1 h1:/tfvTJhfv7hTSL8mWwc5VL4WLLSDL5yn9VqVykdu9r8=
+github.com/openai/openai-go/v2 v2.7.1/go.mod h1:jrJs23apqJKKbT+pqtFgNKpRju/KP9zpUTZhz3GElQE=
 github.com/pelletier/go-toml/v2 v2.2.4 h1:mye9XuhQ6gvn5h28+VilKrrPoQVanw5PMw/TB0t5Ec4=
 github.com/pelletier/go-toml/v2 v2.2.4/go.mod h1:2gIqNv+qfxSVS7cM2xJQKtLSTLUE9V8t9Stt+h56mCY=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=

internal/api/chat/create_conversation_message.go

@@ -180,20 +180,20 @@ func (s *ChatServer) appendConversationMessage(
 
 // 如果 conversationId 是 ""， 就创建新对话，否则就追加消息到对话
 // conversationType 可以在一次 conversation 中多次切换
-func (s *ChatServer) prepare(ctx context.Context, projectId string, conversationId string, userMessage string, userSelectedText string, languageModel models.LanguageModel, conversationType chatv1.ConversationType) (context.Context, *models.Conversation, error) {
+func (s *ChatServer) prepare(ctx context.Context, projectId string, conversationId string, userMessage string, userSelectedText string, languageModel models.LanguageModel, conversationType chatv1.ConversationType) (context.Context, *models.Conversation, *models.Settings, error) {
 	actor, err := contextutil.GetActor(ctx)
 	if err != nil {
-		return ctx, nil, err
+		return ctx, nil, nil, err
 	}
 
 	project, err := s.projectService.GetProject(ctx, actor.ID, projectId)
 	if err != nil && err != mongo.ErrNoDocuments {
-		return ctx, nil, err
+		return ctx, nil, nil, err
 	}
 
 	userInstructions, err := s.userService.GetUserInstructions(ctx, actor.ID)
 	if err != nil {
-		return ctx, nil, err
+		return ctx, nil, nil, err
 	}
 
 	var latexFullSource string
@@ -202,12 +202,12 @@ func (s *ChatServer) prepare(ctx context.Context, projectId string, conversation
 		latexFullSource = "latex_full_source is not available in debug mode"
 	default:
 		if project == nil || project.IsOutOfDate() {
-			return ctx, nil, shared.ErrProjectOutOfDate("project is out of date")
+			return ctx, nil, nil, shared.ErrProjectOutOfDate("project is out of date")
 		}
 
 		latexFullSource, err = project.GetFullContent()
 		if err != nil {
-			return ctx, nil, err
+			return ctx, nil, nil, err
 		}
 	}
 
@@ -238,34 +238,44 @@ func (s *ChatServer) prepare(ctx context.Context, projectId string, conversation
 	}
 
 	if err != nil {
-		return ctx, nil, err
+		return ctx, nil, nil, err
 	}
 
 	ctx = contextutil.SetProjectID(ctx, conversation.ProjectID)
 	ctx = contextutil.SetConversationID(ctx, conversation.ID.Hex())
 
-	return ctx, conversation, nil
+	settings, err := s.userService.GetUserSettings(ctx, actor.ID)
+	if err != nil {
+		return ctx, conversation, nil, err
+	}
+
+	return ctx, conversation, settings, nil
 }
 
 // Deprecated: Use CreateConversationMessageStream instead.
 func (s *ChatServer) CreateConversationMessage(
 	ctx context.Context,
 	req *chatv1.CreateConversationMessageRequest,
 ) (*chatv1.CreateConversationMessageResponse, error) {
-	ctx, conversation, err := s.prepare(
+	languageModel := models.LanguageModel(req.GetLanguageModel())
+	ctx, conversation, settings, err := s.prepare(
 		ctx,
 		req.GetProjectId(),
 		req.GetConversationId(),
 		req.GetUserMessage(),
 		req.GetUserSelectedText(),
-		models.LanguageModel(req.GetLanguageModel()),
+		languageModel,
 		req.GetConversationType(),
 	)
 	if err != nil {
 		return nil, err
 	}
 
-	openaiChatHistory, inappChatHistory, err := s.aiClient.ChatCompletion(ctx, conversation.LanguageModel, conversation.OpenaiChatHistory)
+	llmProvider := &models.LLMProviderConfig{
+		Endpoint: s.cfg.OpenAIBaseURL,
+		APIKey:   settings.OpenAIAPIKey,
+	}
+	openaiChatHistory, inappChatHistory, err := s.aiClient.ChatCompletion(ctx, languageModel, conversation.OpenaiChatHistory, llmProvider)
 	if err != nil {
 		return nil, err
 	}
@@ -290,7 +300,7 @@ func (s *ChatServer) CreateConversationMessage(
 		for i, bsonMsg := range conversation.InappChatHistory {
 			protoMessages[i] = mapper.BSONToChatMessage(bsonMsg)
 		}
-		title, err := s.aiClient.GetConversationTitle(ctx, protoMessages)
+		title, err := s.aiClient.GetConversationTitle(ctx, protoMessages, llmProvider)
 		if err != nil {
 			s.logger.Error("Failed to get conversation title", "error", err, "conversationID", conversation.ID.Hex())
 			return

internal/api/chat/create_conversation_message_stream.go

@@ -24,21 +24,28 @@ func (s *ChatServer) CreateConversationMessageStream(
 	stream chatv1.ChatService_CreateConversationMessageStreamServer,
 ) error {
 	ctx := stream.Context()
-	ctx, conversation, err := s.prepare(
+
+	languageModel := models.LanguageModel(req.GetLanguageModel())
+	ctx, conversation, settings, err := s.prepare(
 		ctx,
 		req.GetProjectId(),
 		req.GetConversationId(),
 		req.GetUserMessage(),
 		req.GetUserSelectedText(),
-		models.LanguageModel(req.GetLanguageModel()),
+		languageModel,
 		req.GetConversationType(),
 	)
 	if err != nil {
 		return s.sendStreamError(stream, err)
 	}
 
 	// 用法跟 ChatCompletion 一样，只是传递了 stream 参数
-	openaiChatHistory, inappChatHistory, err := s.aiClient.ChatCompletionStream(ctx, stream, conversation.ID.Hex(), conversation.LanguageModel, conversation.OpenaiChatHistory)
+	llmProvider := &models.LLMProviderConfig{
+		Endpoint: s.cfg.OpenAIBaseURL,
+		APIKey:   settings.OpenAIAPIKey,
+	}
+
+	openaiChatHistory, inappChatHistory, err := s.aiClient.ChatCompletionStream(ctx, stream, conversation.ID.Hex(), languageModel, conversation.OpenaiChatHistory, llmProvider)
 	if err != nil {
 		return s.sendStreamError(stream, err)
 	}
@@ -64,7 +71,7 @@ func (s *ChatServer) CreateConversationMessageStream(
 			for i, bsonMsg := range conversation.InappChatHistory {
 				protoMessages[i] = mapper.BSONToChatMessage(bsonMsg)
 			}
-			title, err := s.aiClient.GetConversationTitle(ctx, protoMessages)
+			title, err := s.aiClient.GetConversationTitle(ctx, protoMessages, llmProvider)
 			if err != nil {
 				s.logger.Error("Failed to get conversation title", "error", err, "conversationID", conversation.ID.Hex())
 				return

internal/api/chat/list_supported_models.go

@@ -0,0 +1,104 @@
+package chat
+
+import (
+	"context"
+	"strings"
+
+	"paperdebugger/internal/libs/contextutil"
+	chatv1 "paperdebugger/pkg/gen/api/chat/v1"
+
+	"github.com/openai/openai-go/v2"
+)
+
+func (s *ChatServer) ListSupportedModels(
+	ctx context.Context,
+	req *chatv1.ListSupportedModelsRequest,
+) (*chatv1.ListSupportedModelsResponse, error) {
+	actor, err := contextutil.GetActor(ctx)
+	if err != nil {
+		return nil, err
+	}
+
+	settings, err := s.userService.GetUserSettings(ctx, actor.ID)
+	if err != nil {
+		return nil, err
+	}
+
+	var models []*chatv1.SupportedModel
+	if strings.TrimSpace(settings.OpenAIAPIKey) == "" {
+		models = []*chatv1.SupportedModel{
+			{
+
+				Name: "GPT-4o",
+				Slug: openai.ChatModelGPT4o,
+			},
+			{
+				Name: "GPT-4.1",
+				Slug: openai.ChatModelGPT4_1,
+			},
+			{
+				Name: "GPT-4.1-mini",
+				Slug: openai.ChatModelGPT4_1Mini,
+			},
+		}
+	} else {
+		models = []*chatv1.SupportedModel{
+			{
+				Name: "GPT 4o",
+				Slug: openai.ChatModelGPT4o,
+			},
+			{
+				Name: "GPT 4.1",
+				Slug: openai.ChatModelGPT4_1,
+			},
+			{
+				Name: "GPT 4.1 mini",
+				Slug: openai.ChatModelGPT4_1Mini,
+			},
+			{
+				Name: "GPT 5",
+				Slug: openai.ChatModelGPT5,
+			},
+			{
+				Name: "GPT 5 mini",
+				Slug: openai.ChatModelGPT5Mini,
+			},
+			{
+				Name: "GPT 5 nano",
+				Slug: openai.ChatModelGPT5Nano,
+			},
+			{
+				Name: "GPT 5 Chat Latest",
+				Slug: openai.ChatModelGPT5ChatLatest,
+			},
+			{
+				Name: "o1",
+				Slug: openai.ChatModelO1,
+			},
+			{
+				Name: "o1 mini",
+				Slug: openai.ChatModelO1Mini,
+			},
+			{
+				Name: "o3",
+				Slug: openai.ChatModelO3,
+			},
+			{
+				Name: "o3 mini",
+				Slug: openai.ChatModelO3Mini,
+			},
+			{
+				Name: "o4 mini",
+				Slug: openai.ChatModelO4Mini,
+			},
+			{
+				Name: "Codex Mini Latest",
+				Slug: openai.ChatModelCodexMiniLatest,
+			},
+		}
+	}
+
+	return &chatv1.ListSupportedModelsResponse{
+		Models: models,
+	}, nil
+}

internal/api/mapper/user.go

@@ -12,6 +12,7 @@ func MapProtoSettingsToModel(settings *userv1.Settings) *models.Settings {
 		EnableCompletion:             settings.EnableCompletion,
 		FullDocumentRag:              settings.FullDocumentRag,
 		ShowedOnboarding:             settings.ShowedOnboarding,
+		OpenAIAPIKey:                 settings.OpenaiApiKey,
 	}
 }
 
@@ -22,5 +23,6 @@ func MapModelSettingsToProto(settings *models.Settings) *userv1.Settings {
 		EnableCompletion:             settings.EnableCompletion,
 		FullDocumentRag:              settings.FullDocumentRag,
 		ShowedOnboarding:             settings.ShowedOnboarding,
+		OpenaiApiKey:                 settings.OpenAIAPIKey,
 	}
 }

internal/models/language_model.go

@@ -38,6 +38,20 @@ func (x LanguageModel) Name() string {
 		return openai.ChatModelGPT5Mini
 	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_GPT5_NANO:
 		return openai.ChatModelGPT5Nano
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_GPT5_CHAT_LATEST:
+		return openai.ChatModelGPT5ChatLatest
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_O1:
+		return openai.ChatModelO1
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_O1_MINI:
+		return openai.ChatModelO1Mini
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_O3:
+		return openai.ChatModelO3
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_O3_MINI:
+		return openai.ChatModelO3Mini
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_O4_MINI:
+		return openai.ChatModelO4Mini
+	case chatv1.LanguageModel_LANGUAGE_MODEL_OPENAI_CODEX_MINI_LATEST:
+		return openai.ChatModelCodexMiniLatest
 	default:
 		return openai.ChatModelGPT5
 	}

internal/models/llm_provider.go

@@ -0,0 +1,14 @@
+package models
+
+// LLMProviderConfig holds the configuration for LLM API calls.
+// If both Endpoint and APIKey are empty, the system default will be used.
+type LLMProviderConfig struct {
+	Endpoint  string
+	APIKey    string
+	ModelName string
+}
+
+// IsCustom returns true if the user has configured custom LLM provider settings.
+func (c *LLMProviderConfig) IsCustom() bool {
+	return c != nil && c.APIKey != ""
+}

internal/models/user.go

@@ -3,11 +3,12 @@ package models
 import "go.mongodb.org/mongo-driver/v2/bson"
 
 type Settings struct {
-	ShowShortcutsAfterSelection  bool `bson:"show_shortcuts_after_selection"`
-	FullWidthPaperDebuggerButton bool `bson:"full_width_paper_debugger_button"`
-	EnableCompletion             bool `bson:"enable_completion"`
-	FullDocumentRag              bool `bson:"full_document_rag"`
-	ShowedOnboarding             bool `bson:"showed_onboarding"`
+	ShowShortcutsAfterSelection  bool   `bson:"show_shortcuts_after_selection"`
+	FullWidthPaperDebuggerButton bool   `bson:"full_width_paper_debugger_button"`
+	EnableCompletion             bool   `bson:"enable_completion"`
+	FullDocumentRag              bool   `bson:"full_document_rag"`
+	ShowedOnboarding             bool   `bson:"showed_onboarding"`
+	OpenAIAPIKey                 string `bson:"openai_api_key"`
 }
 
 type User struct {

internal/services/toolkit/client/client.go

@@ -17,7 +17,6 @@ import (
 )
 
 type AIClient struct {
-	openaiClient    *openai.Client
 	toolCallHandler *handler.ToolCallHandler
 
 	db                     *mongo.Database
@@ -29,6 +28,29 @@ type AIClient struct {
 	logger                *logger.Logger
 }
 
+// SetOpenAIClient sets the appropriate OpenAI client based on the LLM provider config.
+// If the config specifies a custom endpoint and API key, a new client is created for that endpoint.
+func (a *AIClient) GetOpenAIClient(llmConfig *models.LLMProviderConfig) *openai.Client {
+	var Endpoint string = llmConfig.Endpoint
+	var APIKey string = llmConfig.APIKey
+
+	if Endpoint == "" {
+		Endpoint = a.cfg.OpenAIBaseURL
+	}
+
+	if APIKey == "" {
+		APIKey = a.cfg.OpenAIAPIKey
+	}
+
+	opts := []option.RequestOption{
+		option.WithAPIKey(APIKey),
+		option.WithBaseURL(Endpoint),
+	}
+
+	client := openai.NewClient(opts...)
+	return &client
+}
+
 func NewAIClient(
 	db *db.DB,
 
@@ -73,7 +95,6 @@ func NewAIClient(
 
 	toolCallHandler := handler.NewToolCallHandler(toolRegistry)
 	client := &AIClient{
-		openaiClient:    &oaiClient,
 		toolCallHandler: toolCallHandler,
 
 		db:                     database,