diff --git a/vllm/entrypoints/pooling/pooling/protocol.py b/vllm/entrypoints/pooling/pooling/protocol.py
index b75c34e97902..ab9b132fee34 100644
--- a/vllm/entrypoints/pooling/pooling/protocol.py
+++ b/vllm/entrypoints/pooling/pooling/protocol.py
@@ -97,6 +97,11 @@ def build_tok_params(self, model_config: ModelConfig) -> TokenizeParams:
             max_total_tokens_param="max_model_len",
         )
 
+    def to_pooling_params(self):
+        return PoolingParams(
+            task=self.task,
+        )
+
 
 class IOProcessorResponse(OpenAIBaseModel, Generic[T]):
     request_id: str | None = None