diff --git a/tensorrt_llm/hlapi/llm.py b/tensorrt_llm/hlapi/llm.py
index 7750b6d63..8f389c785 100644
--- a/tensorrt_llm/hlapi/llm.py
+++ b/tensorrt_llm/hlapi/llm.py
@@ -234,7 +234,7 @@ def generate_async(
             prompt = inputs
         elif isinstance(inputs, list) and isinstance(inputs[0], int):
             prompt_token_ids = inputs
-            prompt = None
+            prompt = self.tokenizer.decode(prompt_token_ids)
         else:
             raise TypeError(
                 f"The inputs must be type str or list of int, but got {type(inputs)}"