feat: revamp DeepSeek v4 model handling

- replace legacy DeepSeek ids with the new deepseek-v4 model family\n- move thinking control to request parameters and preserve assistant reasoning content\n- switch history split to IGNORE transcript injection and map upload auth failures to 401\n- update admin defaults, API docs, samples, and tests for the new model scheme

internal/admin/handler_accounts_testing.go

@@ -41,7 +41,7 @@ func (h *Handler) testSingleAccount(w http.ResponseWriter, r *http.Request) {
 	}
 	model, _ := req["model"].(string)
 	if model == "" {
-		model = "deepseek-chat"
+		model = "deepseek-v4-flash"
 	}
 	message, _ := req["message"].(string)
 	result := h.testAccount(r.Context(), acc, model, message)
@@ -53,7 +53,7 @@ func (h *Handler) testAllAccounts(w http.ResponseWriter, r *http.Request) {
 	_ = json.NewDecoder(r.Body).Decode(&req)
 	model, _ := req["model"].(string)
 	if model == "" {
-		model = "deepseek-chat"
+		model = "deepseek-v4-flash"
 	}
 	accounts := h.Store.Snapshot().Accounts
 	if len(accounts) == 0 {
@@ -211,7 +211,7 @@ func (h *Handler) testAPI(w http.ResponseWriter, r *http.Request) {
 	message, _ := req["message"].(string)
 	apiKey, _ := req["api_key"].(string)
 	if model == "" {
-		model = "deepseek-chat"
+		model = "deepseek-v4-flash"
 	}
 	if message == "" {
 		message = "你好"

internal/admin/handler_accounts_testing_test.go

@@ -72,7 +72,7 @@ func TestTestAccount_BatchModeOnlyCreatesSession(t *testing.T) {
 		t.Fatal("expected test account")
 	}
 
-	result := h.testAccount(context.Background(), acc, "deepseek-chat", "")
+	result := h.testAccount(context.Background(), acc, "deepseek-v4-flash", "")
 
 	if ok, _ := result["success"].(bool); !ok {
 		t.Fatalf("expected success=true, got %#v", result)
@@ -177,7 +177,7 @@ func TestTestAccount_MessageModeUsesExpertModelTypeForExpertModel(t *testing.T)
 		t.Fatal("expected test account")
 	}
 
-	result := h.testAccount(context.Background(), acc, "deepseek-expert-chat", "hello")
+	result := h.testAccount(context.Background(), acc, "deepseek-v4-pro", "hello")
 
 	if ok, _ := result["success"].(bool); !ok {
 		t.Fatalf("expected success=true, got %#v", result)
@@ -200,7 +200,7 @@ func TestTestAccount_MessageModeUsesVisionModelTypeForVisionModel(t *testing.T)
 		t.Fatal("expected test account")
 	}
 
-	result := h.testAccount(context.Background(), acc, "deepseek-vision-chat", "hello")
+	result := h.testAccount(context.Background(), acc, "deepseek-v4-vision", "hello")
 
 	if ok, _ := result["success"].(bool); !ok {
 		t.Fatalf("expected success=true, got %#v", result)

internal/admin/handler_chat_history_test.go

@@ -38,7 +38,7 @@ func TestGetChatHistoryAndUpdateSettings(t *testing.T) {
 	entry, err := historyStore.Start(chathistory.StartParams{
 		CallerID:  "caller:test",
 		AccountID: "user@example.com",
-		Model:     "deepseek-chat",
+		Model:     "deepseek-v4-flash",
 		UserInput: "hello",
 	})
 	if err != nil {

internal/admin/handler_raw_samples.go

@@ -114,7 +114,7 @@ func prepareRawSampleCaptureRequest(store ConfigStore, req map[string]any) (map[
 	}
 
 	if model := strings.TrimSpace(fieldString(payload, "model")); model == "" {
-		payload["model"] = "deepseek-chat"
+		payload["model"] = "deepseek-v4-flash"
 	}
 	if _, ok := payload["stream"]; !ok {
 		payload["stream"] = true

internal/admin/handler_raw_samples_test.go

@@ -18,7 +18,7 @@ type stubOpenAIChatCaller struct{}
 
 func (stubOpenAIChatCaller) ChatCompletions(w http.ResponseWriter, _ *http.Request) {
 	store := devcapture.Global()
-	session := store.Start("deepseek_completion", "https://chat.deepseek.com/api/v0/chat/completion", "acct-test", map[string]any{"model": "deepseek-chat"})
+	session := store.Start("deepseek_completion", "https://chat.deepseek.com/api/v0/chat/completion", "acct-test", map[string]any{"model": "deepseek-v4-flash"})
 	raw := io.NopCloser(strings.NewReader(
 		"data: {\"v\":\"hello [reference:1]\"}\n\n" +
 			"data: {\"v\":\"FINISHED\",\"p\":\"response/status\"}\n\n",
@@ -37,7 +37,7 @@ func (stubOpenAIChatCaller) ChatCompletions(w http.ResponseWriter, _ *http.Reque
 type stubOpenAIChatCallerWithContinuations struct{}
 
 func (stubOpenAIChatCallerWithContinuations) ChatCompletions(w http.ResponseWriter, _ *http.Request) {
-	recordCapturedResponse("deepseek_completion", "https://chat.deepseek.com/api/v0/chat/completion", http.StatusOK, map[string]any{"model": "deepseek-chat"}, "data: {\"v\":\"hello [reference:1]\"}\n\n"+"data: [DONE]\n\n")
+	recordCapturedResponse("deepseek_completion", "https://chat.deepseek.com/api/v0/chat/completion", http.StatusOK, map[string]any{"model": "deepseek-v4-flash"}, "data: {\"v\":\"hello [reference:1]\"}\n\n"+"data: [DONE]\n\n")
 	recordCapturedResponse("deepseek_continue", "https://chat.deepseek.com/api/v0/chat/continue", http.StatusOK, map[string]any{"chat_session_id": "session-1", "message_id": 2}, "data: {\"v\":\"continued\"}\n\n"+"data: [DONE]\n\n")
 
 	w.Header().Set("Content-Type", "text/event-stream")
@@ -73,7 +73,7 @@ func TestCaptureRawSampleWritesPersistentSample(t *testing.T) {
 	reqBody := `{
 		"sample_id":"My Sample 01",
 		"api_key":"local-key",
-		"model":"deepseek-chat",
+		"model":"deepseek-v4-flash",
 		"message":"广州天气",
 		"stream":true
 	}`
@@ -130,7 +130,7 @@ func TestCaptureRawSampleCombinesContinuationCaptures(t *testing.T) {
 	reqBody := `{
 		"sample_id":"My Sample 02",
 		"api_key":"local-key",
-		"model":"deepseek-chat",
+		"model":"deepseek-v4-flash",
 		"message":"广州天气",
 		"stream":true
 	}`
@@ -194,13 +194,13 @@ func TestCaptureRawSampleReturnsErrorWhenNoNewCaptureRecorded(t *testing.T) {
 	devcapture.Global().Clear()
 	defer devcapture.Global().Clear()
 
-	recordCapturedResponse("preexisting", "https://chat.deepseek.com/api/v0/chat/completion", http.StatusOK, map[string]any{"model": "deepseek-chat"}, "data: {\"v\":\"old\"}\n\n")
+	recordCapturedResponse("preexisting", "https://chat.deepseek.com/api/v0/chat/completion", http.StatusOK, map[string]any{"model": "deepseek-v4-flash"}, "data: {\"v\":\"old\"}\n\n")
 
 	h := &Handler{OpenAI: stubOpenAIChatCallerWithoutCapture{}}
 	reqBody := `{
 		"sample_id":"My Sample 03",
 		"api_key":"local-key",
-		"model":"deepseek-chat",
+		"model":"deepseek-v4-flash",
 		"message":"广州天气",
 		"stream":true
 	}`

internal/admin/handler_settings_runtime.go

@@ -50,5 +50,5 @@ func settingsClaudeMapping(c config.Config) map[string]string {
 	if len(c.ClaudeModelMap) > 0 {
 		return c.ClaudeModelMap
 	}
-	return map[string]string{"fast": "deepseek-chat", "slow": "deepseek-reasoner"}
+	return map[string]string{"fast": "deepseek-v4-flash", "slow": "deepseek-v4-pro"}
 }