diff --git a/backend/ai_services/views.py b/backend/ai_services/views.py
index 420db9a..9c39fdb 100644
--- a/backend/ai_services/views.py
+++ b/backend/ai_services/views.py
@@ -209,9 +209,26 @@ class TranscriptionTaskViewSet(viewsets.ModelViewSet):
                     except Exception as e:
                         logger.error(f"Error downloading transcription: {e}")
                         transcription_data = {}
-
+                elif isinstance(transcription_data, dict) and 'TranscriptionUrl' in transcription_data:
+                    # 有些情况 Transcription 还是对象，但内容在 Url 字段
+                     try:
+                        import requests
+                        url = transcription_data['TranscriptionUrl']
+                        logger.info(f"Downloading transcription from {url}")
+                        t_resp = requests.get(url)
+                        if t_resp.status_code == 200:
+                            transcription_data = t_resp.json()
+                     except Exception as e:
+                         logger.error(f"Error downloading transcription nested url: {e}")
+                
                 if isinstance(transcription_data, dict):
+                    # 尝试多种可能的路径提取句子
+                    # 1. 直接在根目录: {"Sentences": [...]}
+                    # 2. 在 Transcription 字段下: {"Transcription": {"Sentences": [...]}}
                     sentences = transcription_data.get('Sentences', [])
+                    if not sentences and 'Transcription' in transcription_data:
+                         sentences = transcription_data['Transcription'].get('Sentences', [])
+                    
                     full_text = " ".join([s.get('Text', '') for s in sentences])
                     task.transcription = full_text