diff --git a/src/transformers/models/llava_next/processing_llava_next.py b/src/transformers/models/llava_next/processing_llava_next.py
index db9d1276832b39..38173cbd861fc1 100644
--- a/src/transformers/models/llava_next/processing_llava_next.py
+++ b/src/transformers/models/llava_next/processing_llava_next.py
@@ -165,7 +165,8 @@ def __call__(
                         image_size = next(image_sizes)
                         if not isinstance(image_size, (list, tuple)):
                             # cast to list to avoid numerical precision errors when calculating unpadding
-                            orig_height, orig_width = image_size.tolist()
+                            image_size = image_size.tolist()
+                        orig_height, orig_width = image_size
                         num_image_tokens = self._get_number_of_features(orig_height, orig_width, height, width)
                         if self.vision_feature_select_strategy == "default":
                             num_image_tokens -= self.num_additional_image_tokens
diff --git a/src/transformers/models/llava_next_video/processing_llava_next_video.py b/src/transformers/models/llava_next_video/processing_llava_next_video.py
index 5805782f779e5f..65195b77240721 100644
--- a/src/transformers/models/llava_next_video/processing_llava_next_video.py
+++ b/src/transformers/models/llava_next_video/processing_llava_next_video.py
@@ -192,7 +192,8 @@ def __call__(
                         image_size = next(image_sizes)
                         if not isinstance(image_size, (list, tuple)):
                             # cast to list to avoid numerical precision errors when calculating unpadding
-                            orig_height, orig_width = image_size.tolist()
+                            image_size = image_size.tolist()
+                        orig_height, orig_width = image_size
                         num_image_tokens = self._get_number_of_features(orig_height, orig_width, height, width)
                         if self.vision_feature_select_strategy == "default":
                             num_image_tokens -= self.num_additional_image_tokens
diff --git a/src/transformers/models/llava_onevision/processing_llava_onevision.py b/src/transformers/models/llava_onevision/processing_llava_onevision.py
index 4f67f9e4c030e8..ff808802bcaa63 100644
--- a/src/transformers/models/llava_onevision/processing_llava_onevision.py
+++ b/src/transformers/models/llava_onevision/processing_llava_onevision.py
@@ -191,7 +191,8 @@ def _expand_image_tokens(
                 original_size = image_size_list[0] if num_frames != 1 else image_size_list
                 if not isinstance(original_size, (list, tuple)):
                     # cast to list to avoid numerical precision errors when calculating unpadding
-                    orig_height, orig_width = original_size.tolist()
+                    original_size = original_size.tolist()
+                orig_height, orig_width = original_size
                 num_image_tokens = self._get_number_of_features(orig_height, orig_width, height, width)
                 if self.vision_feature_select_strategy == "default":
                     num_image_tokens -= 1