fix: 修复返回给前端的 totalTokens 为0的问题并提供示例 .env

backend/internal/orchestrator/pipeline.go

@@ -166,11 +166,12 @@ func (p *Pipeline) ProcessQuery(
 	var ttsErr error
 
 	// goroutine 1: 消费 LLM token + 句子切分
+	var tokenUsage *llm.TokenUsage
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
 		defer close(sentenceCh)
-		fullText = p.consumeLLMStream(ctx, llmStream, req.RequestID, sender, splitter)
+		fullText, tokenUsage = p.consumeLLMStream(ctx, llmStream, req.RequestID, sender, splitter)
 	}()
 
 	// goroutine 2: TTS 合成（如果启用）
@@ -205,13 +206,25 @@ func (p *Pipeline) ProcessQuery(
 
 	// 发送 llm_done
 	latency := time.Since(startTime).Milliseconds()
-	if err := sender.SendLLMDone(models.WsLLMDone{
+	done := models.WsLLMDone{
 		Type:      "llm_done",
 		RequestID: req.RequestID,
 		FullText:  fullText,
 		Model:     p.model,
 		LatencyMs: latency,
-	}); err != nil {
+	}
+	if tokenUsage != nil {
+		done.TokensUsed = struct {
+			Prompt     int `json:"prompt"`
+			Completion int `json:"completion"`
+			Total      int `json:"total"`
+		}{
+			Prompt:     tokenUsage.Prompt,
+			Completion: tokenUsage.Completion,
+			Total:      tokenUsage.Total,
+		}
+	}
+	if err := sender.SendLLMDone(done); err != nil {
 		log.Errorw("发送 llm_done 失败", "error", err)
 	}
 
@@ -225,33 +238,36 @@ func (p *Pipeline) ProcessQuery(
 }
 
 // consumeLLMStream 消费 LLM 流式输出，发送 llm_chunk 并进行句子切分。
+// 返回完整文本和 token 用量。
 func (p *Pipeline) consumeLLMStream(
 	ctx context.Context,
 	stream <-chan llm.Chunk,
 	requestID string,
 	sender Sender,
 	splitter *Splitter,
-) string {
+) (string, *llm.TokenUsage) {
 	log := logger.Log
 	var fullText strings.Builder
+	var tokenUsage *llm.TokenUsage
 
 	for chunk := range stream {
 		// 检查上下文是否已取消
 		select {
 		case <-ctx.Done():
 			log.Infow("LLM 流被中断", "request_id", requestID)
-			return fullText.String()
+			return fullText.String(), tokenUsage
 		default:
 		}
 
 		if chunk.Done {
-			// 流结束
+			// 流结束，记录 token 用量
 			if chunk.TokensUsed != nil {
+				tokenUsage = chunk.TokensUsed
 				log.Infow("LLM 用量统计",
 					"request_id", requestID,
-					"prompt_tokens", chunk.TokensUsed.Prompt,
-					"completion_tokens", chunk.TokensUsed.Completion,
-					"total_tokens", chunk.TokensUsed.Total,
+					"prompt_tokens", tokenUsage.Prompt,
+					"completion_tokens", tokenUsage.Completion,
+					"total_tokens", tokenUsage.Total,
 				)
 			}
 			break
@@ -277,7 +293,7 @@ func (p *Pipeline) consumeLLMStream(
 	// 刷新切分器中的剩余文本
 	splitter.Flush()
 
-	return fullText.String()
+	return fullText.String(), tokenUsage
 }
 
 // synthesizeTTS 从句子 channel 读取文本，进行 TTS 合成并发送音频。