diff --git a/backend/controllers/ai.go b/backend/controllers/ai.go
index 7626075..8b8deec 100644
--- a/backend/controllers/ai.go
+++ b/backend/controllers/ai.go
@@ -64,12 +64,12 @@ func (ctrl *AIController) TestConfig(c *gin.Context) {
 		cfg.APIKey = req.APIKey
 	}
 
-	reply, err := services.TestAIConnection(cfg)
+	result, err := services.TestAIConnection(cfg)
 	if err != nil {
 		utils.Error(c, 400, err.Error())
 		return
 	}
-	utils.Success(c, gin.H{"reply": reply, "model": cfg.Model})
+	utils.Success(c, result)
 }
 
 func containsMask(s string) bool {
diff --git a/backend/services/ai_provider.go b/backend/services/ai_provider.go
index 9d5bd24..03f0fc8 100644
--- a/backend/services/ai_provider.go
+++ b/backend/services/ai_provider.go
@@ -115,17 +115,36 @@ func aiChat(ctx context.Context, cfg models.AIConfig, messages []aiMessage) (str
 	return parsed.Choices[0].Message.Content, parsed.Usage.PromptTokens, parsed.Usage.CompletionTokens, nil
 }
 
-// TestAIConnection 连通性测试：发一句最小请求，验证 baseURL/key/model 三件套。
-func TestAIConnection(cfg models.AIConfig) (string, error) {
+// AIConnectionTest 连通性测试结果:面向展示的结构化信息,不外发原始回复
+type AIConnectionTest struct {
+	Model       string `json:"model"`
+	LatencyMs   int64  `json:"latencyMs"`
+	TokensUsed  int    `json:"tokensUsed"`
+	ReplyLength int    `json:"replyLength"`
+}
+
+// TestAIConnection 连通性测试:发一句最小请求,验证 baseURL/key/model 三件套。
+// 返回结构化结果(模型/耗时/token);模型原始回复不回传 —— 测试要的是"通没通、多快",不是看模型说了什么。
+func TestAIConnection(cfg models.AIConfig) (*AIConnectionTest, error) {
 	cfg.MaxTokens = 32
 	cfg.TimeoutSec = 30
-	reply, _, _, err := aiChat(context.Background(), cfg, []aiMessage{
+	start := time.Now()
+	reply, promptTokens, completionTokens, err := aiChat(context.Background(), cfg, []aiMessage{
 		{Role: "user", Content: "只回复两个字：正常"},
 	})
 	if err != nil {
-		return "", err
+		return nil, err
 	}
-	return strings.TrimSpace(reply), nil
+	trimmed := strings.TrimSpace(reply)
+	if trimmed == "" {
+		return nil, fmt.Errorf("模型返回了空内容(检查 baseURL 是否为 OpenAI 兼容端点)")
+	}
+	return &AIConnectionTest{
+		Model:       cfg.Model,
+		LatencyMs:   time.Since(start).Milliseconds(),
+		TokensUsed:  promptTokens + completionTokens,
+		ReplyLength: len([]rune(trimmed)),
+	}, nil
 }
 
 // ChatMessage 浮窗助手的多轮消息(user/assistant 交替)
diff --git a/frontend/src/api/ai/index.ts b/frontend/src/api/ai/index.ts
index 71be0d2..c65849f 100644
--- a/frontend/src/api/ai/index.ts
+++ b/frontend/src/api/ai/index.ts
@@ -83,7 +83,7 @@ export const aiApi = {
   getConfig: () => http.get<AIConfigResponse>('/ai/config'),
   saveConfig: (data: Partial<AIConfig>) => http.put<{ config: AIConfig }>('/ai/config', data),
   testConfig: (data: { baseUrl?: string; model?: string; apiKey?: string }) =>
-    http.post<{ reply: string; model: string }>('/ai/config/test', data, { timeout: 45000 }),
+    http.post<{ model: string; latencyMs: number; tokensUsed: number; replyLength: number }>('/ai/config/test', data, { timeout: 45000 }),
   analyze: (data: { sourceId: number; start: number; end: number; question?: string; app?: string }) =>
     http.post<AIAnalyzeResponse>('/ai/analyze', data, { timeout: 180000 }),
   runs: (page: number, pageSize: number) =>
diff --git a/frontend/src/i18n/locales/en/aiops.ts b/frontend/src/i18n/locales/en/aiops.ts
index 1262ae3..38ae7ea 100644
--- a/frontend/src/i18n/locales/en/aiops.ts
+++ b/frontend/src/i18n/locales/en/aiops.ts
@@ -93,7 +93,7 @@ export default {
     save: 'Save',
     test: 'Test connection',
     testing: 'Testing…',
-    testOk: 'Connection OK, model replied: {{reply}}',
+    testOk: 'Connected: {{model}} · {{latency}}s · {{tokens}} tokens',
     usedToday: '{{used}} tokens used today',
   },
 
diff --git a/frontend/src/i18n/locales/zh/aiops.ts b/frontend/src/i18n/locales/zh/aiops.ts
index 2c972a6..349c76e 100644
--- a/frontend/src/i18n/locales/zh/aiops.ts
+++ b/frontend/src/i18n/locales/zh/aiops.ts
@@ -93,7 +93,7 @@ export default {
     save: '保存配置',
     test: '测试连接',
     testing: '测试中…',
-    testOk: '连接正常，模型回复：{{reply}}',
+    testOk: '连接正常:{{model}} · {{latency}}s · {{tokens}} tokens',
     usedToday: '今日已用 {{used}} tokens',
   },
 
diff --git a/frontend/src/pages/aiops/index.tsx b/frontend/src/pages/aiops/index.tsx
index da1e950..8316bc4 100644
--- a/frontend/src/pages/aiops/index.tsx
+++ b/frontend/src/pages/aiops/index.tsx
@@ -207,7 +207,7 @@ export default function AiopsPage() {
         model: config?.model,
         apiKey: apiKeyInput.trim() || undefined,
       })
-      toast.success(t('config.testOk', { reply: res.reply || 'OK' }))
+      toast.success(t('config.testOk', { model: res.model, latency: (res.latencyMs / 1000).toFixed(1), tokens: res.tokensUsed }))
     } catch (err) {
       toast.error(err instanceof Error ? err.message : '连接失败')
     } finally {