diff --git a/main/xiaozhi-server/core/providers/asr/aliyun_stream.py b/main/xiaozhi-server/core/providers/asr/aliyun_stream.py
index a5f327c0..2f16a975 100644
--- a/main/xiaozhi-server/core/providers/asr/aliyun_stream.py
+++ b/main/xiaozhi-server/core/providers/asr/aliyun_stream.py
@@ -120,6 +120,14 @@ class ASRProvider(ASRProviderBase):
         await super().open_audio_channels(conn)
 
     async def receive_audio(self, conn, audio, audio_have_voice):
+        # 初始化音频缓存
+        if not hasattr(conn, 'asr_audio_for_voiceprint'):
+            conn.asr_audio_for_voiceprint = []
+        
+        # 存储音频数据
+        if audio:
+            conn.asr_audio_for_voiceprint.append(audio)
+        
         conn.asr_audio.append(audio)
         conn.asr_audio = conn.asr_audio[-10:]
 
@@ -129,7 +137,7 @@ class ASRProvider(ASRProviderBase):
                 await self._start_recognition(conn)
             except Exception as e:
                 logger.bind(tag=TAG).error(f"开始识别失败: {str(e)}")
-                await self._cleanup()
+                await self._cleanup(conn)
                 return
 
         if self.asr_ws and self.is_processing and self.server_ready:
@@ -138,7 +146,7 @@ class ASRProvider(ASRProviderBase):
                 await self.asr_ws.send(pcm_frame)
             except Exception as e:
                 logger.bind(tag=TAG).warning(f"发送音频失败: {str(e)}")
-                await self._cleanup()
+                await self._cleanup(conn)
 
     async def _start_recognition(self, conn):
         """开始识别会话"""
@@ -235,7 +243,11 @@ class ASRProvider(ASRProviderBase):
                         if text:
                             self.text = text
                             conn.reset_vad_states()
-                            await self.handle_voice_stop(conn, None)
+                            # 传递缓存的音频数据
+                            audio_data = getattr(conn, 'asr_audio_for_voiceprint', [])
+                            await self.handle_voice_stop(conn, audio_data)
+                            # 清空缓存
+                            conn.asr_audio_for_voiceprint = []
                             break
                     elif message_name == "TranscriptionCompleted":
                         # 识别完成
@@ -253,12 +265,16 @@ class ASRProvider(ASRProviderBase):
         except Exception as e:
             logger.bind(tag=TAG).error(f"结果转发失败: {str(e)}")
         finally:
-            await self._cleanup()
+            await self._cleanup(conn)
 
-    async def _cleanup(self):
+    async def _cleanup(self, conn):
         """清理资源"""
         logger.bind(tag=TAG).info(f"开始ASR会话清理 | 当前状态: processing={self.is_processing}, server_ready={self.server_ready}")
         
+        # 清理连接的音频缓存
+        if conn and hasattr(conn, 'asr_audio_for_voiceprint'):
+            conn.asr_audio_for_voiceprint = []
+        
         # 判断是否需要发送终止请求
         should_stop = self.is_processing or self.server_ready