diff --git a/app/services/qa_service.py b/app/services/qa_service.py
index a8136b7..fe23c80 100644
--- a/app/services/qa_service.py
+++ b/app/services/qa_service.py
@@ -3,6 +3,7 @@ import base64
 from app.clients.llm.base import LLMClient
 from app.clients.storage.base import StorageClient
 from app.core.config import get_config
+from app.core.exceptions import LLMParseError
 from app.core.json_utils import extract_json
 from app.core.logging import get_logger
 from app.models.qa_models import (
@@ -47,10 +48,15 @@ async def gen_text_qa(req: GenTextQARequest, llm: LLMClient) -> TextQAResponse:
     messages = [{"role": "user", "content": prompt}]
     raw = await llm.chat(model, messages)
 
+    items_raw = extract_json(raw)
     logger.info("gen_text_qa", extra={"items": len(req.items), "model": model})
 
-    items_raw = extract_json(raw)
-    pairs = [QAPair(question=item["question"], answer=item["answer"]) for item in items_raw]
+    if not isinstance(items_raw, list):
+        raise LLMParseError("大模型返回的问答对格式不正确")
+    try:
+        pairs = [QAPair(question=item["question"], answer=item["answer"]) for item in items_raw]
+    except (KeyError, TypeError):
+        raise LLMParseError("大模型返回的问答对格式不正确")
     return TextQAResponse(pairs=pairs)
 
 
@@ -91,16 +97,22 @@ async def gen_image_qa(
         ]
 
         raw = await llm.chat_vision(model, messages)
-        logger.info("gen_image_qa", extra={"path": item.cropped_image_path, "model": model})
 
         items_raw = extract_json(raw)
-        for qa in items_raw:
-            pairs.append(
-                ImageQAPair(
-                    question=qa["question"],
-                    answer=qa["answer"],
-                    image_path=item.cropped_image_path,
+        logger.info("gen_image_qa", extra={"path": item.cropped_image_path, "model": model})
+
+        if not isinstance(items_raw, list):
+            raise LLMParseError("大模型返回的问答对格式不正确")
+        try:
+            for qa in items_raw:
+                pairs.append(
+                    ImageQAPair(
+                        question=qa["question"],
+                        answer=qa["answer"],
+                        image_path=item.cropped_image_path,
+                    )
                 )
-            )
+        except (KeyError, TypeError):
+            raise LLMParseError("大模型返回的问答对格式不正确")
 
     return ImageQAResponse(pairs=pairs)