diff --git a/src/transformers/modeling_utils.py b/src/transformers/modeling_utils.py
index a11f7743ed8e..33ecceacb17a 100644
--- a/src/transformers/modeling_utils.py
+++ b/src/transformers/modeling_utils.py
@@ -4015,8 +4015,11 @@ def save_pretrained(
             repo_id = self._create_repo(repo_id, **kwargs)
             files_timestamps = self._get_files_timestamps(save_directory)
 
+        metadata = {}
         if hf_quantizer is not None:
-            state_dict = hf_quantizer.get_state_dict(self)
+            state_dict, metadata = hf_quantizer.get_state_dict_and_metadata(self, safe_serialization)
+        metadata["format"] = "pt"
+
         # Only save the model itself if we are using distributed training
         model_to_save = unwrap_model(self)
         # save the string version of dtype to the config, e.g. convert torch.float32 => "float32"
@@ -4294,7 +4297,7 @@ def save_pretrained(
             if safe_serialization:
                 # At some point we will need to deal better with save_function (used for TPU and other distributed
                 # joyfulness), but for now this enough.
-                safe_save_file(shard, os.path.join(save_directory, shard_file), metadata={"format": "pt"})
+                safe_save_file(shard, os.path.join(save_directory, shard_file), metadata=metadata)
             else:
                 save_function(shard, os.path.join(save_directory, shard_file))
 
diff --git a/src/transformers/quantizers/base.py b/src/transformers/quantizers/base.py
index 653953abec0a..323faa9c17e2 100644
--- a/src/transformers/quantizers/base.py
+++ b/src/transformers/quantizers/base.py
@@ -338,9 +338,9 @@ def is_compileable(self) -> bool:
         """Flag indicating whether the quantized model can be compiled"""
         return False
 
-    def get_state_dict(self, model):
-        """Get state dict. Useful when we need to modify a bit the state dict due to quantization"""
-        return None
+    def get_state_dict_and_metadata(self, model, safe_serialization=False):
+        """Get state dict and metadata. Useful when we need to modify a bit the state dict due to quantization"""
+        return None, {}
 
     @abstractmethod
     def _process_model_before_weight_loading(self, model, **kwargs): ...
diff --git a/src/transformers/quantizers/quantizer_mxfp4.py b/src/transformers/quantizers/quantizer_mxfp4.py
index b9076007d38d..d0d370a11df6 100644
--- a/src/transformers/quantizers/quantizer_mxfp4.py
+++ b/src/transformers/quantizers/quantizer_mxfp4.py
@@ -379,7 +379,7 @@ def update_param_name(self, param_name: str) -> str:
                 return param_name.replace("down_proj", "down_proj_blocks")
         return param_name
 
-    def get_state_dict(self, model):
+    def get_state_dict_and_metadata(self, model):
         from ..integrations import Mxfp4GptOssExperts
 
         state_dict = model.state_dict()
@@ -411,7 +411,8 @@ def get_state_dict(self, model):
                     ).transpose(-1, -2)
                 )
 
-        return state_dict
+        metadata = {}
+        return state_dict, metadata
 
     def is_serializable(self, safe_serialization=None):
         return True