diff --git a/app/chain/subscribe.py b/app/chain/subscribe.py
index 7cceffb2..3b68c8e6 100644
--- a/app/chain/subscribe.py
+++ b/app/chain/subscribe.py
@@ -801,15 +801,27 @@ class SubscribeChain(ChainBase):
                 for context in contexts:
                     if global_vars.is_system_stopped:
                         break
-                    # 如果种子未识别，尝试识别
-                    if not context.media_info or (not context.media_info.tmdb_id
-                                                  and not context.media_info.douban_id):
+                    # 如果种子未识别且失败次数未超过3次，尝试识别
+                    if (not context.media_info or (not context.media_info.tmdb_id
+                                                   and not context.media_info.douban_id)) and context.media_recognize_fail_count < 3:
+                        logger.debug(
+                            f'尝试重新识别种子：{context.torrent_info.title}，当前失败次数：{context.media_recognize_fail_count}/3')
                         re_mediainfo = self.recognize_media(meta=context.meta_info)
                         if re_mediainfo:
                             # 清理多余信息
                             re_mediainfo.clear()
                             # 更新种子缓存
                             context.media_info = re_mediainfo
+                            # 重置失败次数
+                            context.media_recognize_fail_count = 0
+                            logger.debug(f'种子 {context.torrent_info.title} 重新识别成功')
+                        else:
+                            # 识别失败，增加失败次数
+                            context.media_recognize_fail_count += 1
+                            logger.debug(
+                                f'种子 {context.torrent_info.title} 媒体识别失败，失败次数：{context.media_recognize_fail_count}/3')
+                    elif context.media_recognize_fail_count >= 3:
+                        logger.debug(f'种子 {context.torrent_info.title} 已达到最大识别失败次数(3次)，跳过识别')
                     # 添加已预处理
                     processed_torrents[domain].append(context)
 
diff --git a/app/chain/torrents.py b/app/chain/torrents.py
index 78d8443c..f15c9f58 100644
--- a/app/chain/torrents.py
+++ b/app/chain/torrents.py
@@ -56,9 +56,14 @@ class TorrentsChain(ChainBase):
 
         # 读取缓存
         if stype == 'spider':
-            return self.load_cache(self._spider_file) or {}
+            torrents_cache = self.load_cache(self._spider_file) or {}
         else:
-            return self.load_cache(self._rss_file) or {}
+            torrents_cache = self.load_cache(self._rss_file) or {}
+
+        # 兼容性处理：为旧版本的Context对象添加失败次数字段
+        self._ensure_context_compatibility(torrents_cache)
+
+        return torrents_cache
 
     async def async_get_torrents(self, stype: Optional[str] = None) -> Dict[str, List[Context]]:
         """
@@ -71,9 +76,14 @@ class TorrentsChain(ChainBase):
 
         # 异步读取缓存
         if stype == 'spider':
-            return await self.async_load_cache(self._spider_file) or {}
+            torrents_cache = await self.async_load_cache(self._spider_file) or {}
         else:
-            return await self.async_load_cache(self._rss_file) or {}
+            torrents_cache = await self.async_load_cache(self._rss_file) or {}
+
+        # 兼容性处理：为旧版本的Context对象添加失败次数字段
+        self._ensure_context_compatibility(torrents_cache)
+
+        return torrents_cache
 
     def clear_torrents(self):
         """
@@ -274,6 +284,9 @@ class TorrentsChain(ChainBase):
                         mediainfo.clear()
                         # 上下文
                         context = Context(meta_info=meta, media_info=mediainfo, torrent_info=torrent)
+                        # 如果未识别到媒体信息，设置初始失败次数为1
+                        if not mediainfo or (not mediainfo.tmdb_id and not mediainfo.douban_id):
+                            context.media_recognize_fail_count = 1
                         # 添加到缓存
                         if not torrents_cache.get(domain):
                             torrents_cache[domain] = [context]
@@ -300,6 +313,21 @@ class TorrentsChain(ChainBase):
 
         return torrents_cache
 
+    @staticmethod
+    def _ensure_context_compatibility(torrents_cache: Dict[str, List[Context]]):
+        """
+        确保Context对象的兼容性，为旧版本添加缺失的字段
+        """
+        for domain, contexts in torrents_cache.items():
+            for context in contexts:
+                # 如果Context对象没有media_recognize_fail_count字段，添加默认值
+                if not hasattr(context, 'media_recognize_fail_count'):
+                    context.media_recognize_fail_count = 0
+                    # 如果媒体信息未识别，设置初始失败次数
+                    if (not context.media_info or
+                            (not context.media_info.tmdb_id and not context.media_info.douban_id)):
+                        context.media_recognize_fail_count = 1
+
     def __renew_rss_url(self, domain: str, site: dict):
         """
         保留原配置生成新的rss地址
diff --git a/app/core/context.py b/app/core/context.py
index d5b26f5e..9beaff55 100644
--- a/app/core/context.py
+++ b/app/core/context.py
@@ -814,6 +814,8 @@ class Context:
     media_info: MediaInfo = None
     # 种子信息
     torrent_info: TorrentInfo = None
+    # 媒体识别失败次数
+    media_recognize_fail_count: int = 0
 
     def to_dict(self):
         """
@@ -822,5 +824,6 @@ class Context:
         return {
             "meta_info": self.meta_info.to_dict() if self.meta_info else None,
             "torrent_info": self.torrent_info.to_dict() if self.torrent_info else None,
-            "media_info": self.media_info.to_dict() if self.media_info else None
+            "media_info": self.media_info.to_dict() if self.media_info else None,
+            "media_recognize_fail_count": self.media_recognize_fail_count
         }