diff --git a/vllm/model_executor/models/gemma3_mm.py b/vllm/model_executor/models/gemma3_mm.py
index 619d2aa67491..dea45d0eb67a 100644
--- a/vllm/model_executor/models/gemma3_mm.py
+++ b/vllm/model_executor/models/gemma3_mm.py
@@ -479,8 +479,12 @@ class Gemma3ForConditionalGeneration(nn.Module, SupportsMultiModal, SupportsPP,
             "model.vision_tower.": "vision_tower.",
             "model.multi_modal_projector.": "multi_modal_projector.",
             "lm_head.": "language_model.lm_head.",
-            "vision_tower.vision_model.": "vision_model.",
-        })
+        },
+        orig_to_new_substr={
+            # mapping for llm-compressor quantized checkpoints
+            "vision_model.vision_model.": "vision_model.",
+        },
+    )
 
     def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         super().__init__()