diff --git a/vllm_omni/worker/gpu_model_runner.py b/vllm_omni/worker/gpu_model_runner.py
index 1f678b579fa..5ff62c11b40 100644
--- a/vllm_omni/worker/gpu_model_runner.py
+++ b/vllm_omni/worker/gpu_model_runner.py
@@ -1241,6 +1241,7 @@ def _preprocess(
                 span_len = int(e) - int(s)
 
                 # call the custom process function
+                req_infos["request_id"] = req_id
                 embed_slice = inputs_embeds[s:e] if inputs_embeds is not None else None
                 req_input_ids, req_embeds, update_dict = self.model.preprocess(
                     input_ids=input_ids[s:e], input_embeds=embed_slice, **req_infos