diff --git a/vllm_ascend/patch/worker/__init__.py b/vllm_ascend/patch/worker/__init__.py
index e9bf8b0d17a..537752e875e 100644
--- a/vllm_ascend/patch/worker/__init__.py
+++ b/vllm_ascend/patch/worker/__init__.py
@@ -29,6 +29,7 @@
 import vllm_ascend.patch.worker.patch_deepseek  # noqa
 import vllm_ascend.patch.worker.patch_multimodal_merge  # noqa
 import vllm_ascend.patch.worker.patch_minicpm  # noqa
+import vllm_ascend.patch.worker.patch_model_runner  # noqa
 import vllm_ascend.patch.worker.patch_rope  # noqa
 import vllm_ascend.patch.worker.patch_qwen3_next  # noqa
 import vllm_ascend.patch.worker.patch_qwen3_next_mtp  # noqa
diff --git a/vllm_ascend/platform.py b/vllm_ascend/platform.py
index 21746346c19..db9e7c83147 100644
--- a/vllm_ascend/platform.py
+++ b/vllm_ascend/platform.py
@@ -333,12 +333,6 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
                 vllm_config)
             vllm_config.scheduler_config = recompute_scheduler_config
 
-            # When open PD separation + MTP + Full Graph + asynchronous scheduling,
-            # patch gpu_model_runner._update_states for solving
-            # decode nodes may be in eager mode.
-            if vllm_config.speculative_config and vllm_config.kv_transfer_config:
-                import vllm_ascend.patch.worker.patch_model_runner  # noqa
-
         # Extend original scheduler_config to use SchedulerDynamicBatch.
         if ascend_config.SLO_limits_for_dynamic_batch != -1:
             vllm_config.scheduler_config.scheduler_cls = (