diff --git a/docs/pipeline_usage/tutorials/ocr_pipelines/PaddleOCR-VL.md b/docs/pipeline_usage/tutorials/ocr_pipelines/PaddleOCR-VL.md
index 455ccd09d1..251deaa29c 100644
--- a/docs/pipeline_usage/tutorials/ocr_pipelines/PaddleOCR-VL.md
+++ b/docs/pipeline_usage/tutorials/ocr_pipelines/PaddleOCR-VL.md
@@ -984,6 +984,7 @@ VLRecognition:
   genai_config:
     backend: vllm-server
     server_url: http://127.0.0.1:8118/v1
+    api_key: "secret"
 ```
 
 之后，可以使用修改好的配置文件进行产线调用。例如通过 CLI 调用：
diff --git a/paddlex/inference/models/base/predictor/base_predictor.py b/paddlex/inference/models/base/predictor/base_predictor.py
index db0a4323d0..4ef3262224 100644
--- a/paddlex/inference/models/base/predictor/base_predictor.py
+++ b/paddlex/inference/models/base/predictor/base_predictor.py
@@ -136,6 +136,7 @@ def __init__(
             self._genai_client = GenAIClient(
                 backend=genai_config.backend,
                 base_url=genai_config.server_url,
+                api_key=genai_config.api_key,
                 max_concurrency=genai_config.max_concurrency,
                 model_name=model_name,
                 **(genai_config.client_kwargs or {}),
diff --git a/paddlex/inference/models/common/genai.py b/paddlex/inference/models/common/genai.py
index ce274dd394..b80675667f 100644
--- a/paddlex/inference/models/common/genai.py
+++ b/paddlex/inference/models/common/genai.py
@@ -33,6 +33,7 @@ class GenAIConfig(BaseModel):
         "native"
     )
     server_url: Optional[str] = None
+    api_key: Optional[str] = "null"
     max_concurrency: int = 200
     client_kwargs: Optional[Dict[str, Any]] = None