diff --git a/pipeline_service/llm_bridge.py b/pipeline_service/llm_bridge.py
index e7ee630..70b405e 100644
--- a/pipeline_service/llm_bridge.py
+++ b/pipeline_service/llm_bridge.py
@@ -105,7 +105,8 @@ def _no_llm_error(model_name=None, org_id=None) -> ValueError:
 
 
 async def llm_call(prompt: str, model: str = None, temperature: float = 0.7,
-                   org_id: str = None, user_id: str = None, purpose: str = '') -> str:
+                   org_id: str = None, user_id: str = None, purpose: str = '',
+                   timeout: int = 0) -> str:
     """Call LLM and return text response.
 
     统一走模型治理推理 API（门禁链 + 双维度记账）。
@@ -133,6 +134,8 @@ async def llm_call(prompt: str, model: str = None, temperature: float = 0.7,
     }
     if purpose:
         payload["_purpose"] = purpose
+    if timeout:
+        payload["_timeout"] = int(timeout)
     data = await _http_chat(payload, org_id or '0', user_id or '', model or '')
     try:
         return data["choices"][0]["message"]["content"]
@@ -147,14 +150,18 @@ async def call_llm(tenant_id: str, prompt: str, model: str = None, temperature:
 
 
 async def llm_call_msgs(messages: list, model: str = None, temperature: float = 0.7,
-                        org_id: str = None, user_id: str = None, purpose: str = '') -> str:
+                        org_id: str = None, user_id: str = None, purpose: str = '',
+                        timeout: int = 0) -> str:
     """Call LLM with full message array (system/user/assistant).
 
     purpose='utility'：辅助任务（分类/选择/摘要），治理层按用途选模型链。
+    timeout：单次上游调用超时秒数（0=用端点默认；上限 900，超长文本提取用）。
     """
     payload = {"model": model or '', "messages": messages, "temperature": temperature}
     if purpose:
         payload["_purpose"] = purpose
+    if timeout:
+        payload["_timeout"] = int(timeout)
     data = await _http_chat(payload, org_id or '0', user_id or '', model or '')
     try:
         return data["choices"][0]["message"]["content"]