diff --git a/vllm/model_executor/model_loader/loader.py b/vllm/model_executor/model_loader/loader.py
index 2cb9e0ca7c505..2cf4e92908353 100644
--- a/vllm/model_executor/model_loader/loader.py
+++ b/vllm/model_executor/model_loader/loader.py
@@ -150,7 +150,7 @@ def _get_model_initialization_kwargs(
 
 
 def build_model(model_class: Type[nn.Module],
-                vllm_config: VllmConfig,
+                vllm_config: Optional[VllmConfig],
                 hf_config: PretrainedConfig,
                 cache_config: Optional[CacheConfig],
                 quant_config: Optional[QuantizationConfig],
diff --git a/vllm/model_executor/models/utils.py b/vllm/model_executor/models/utils.py
index c6ec1769fc5d1..fee97e8922a76 100644
--- a/vllm/model_executor/models/utils.py
+++ b/vllm/model_executor/models/utils.py
@@ -252,6 +252,7 @@ def init_vllm_registered_model(
 
     return build_model(
         model_class,
+        None,
         hf_config,
         cache_config,
         quant_config,