diff --git a/tests/models/multimodal/generation/test_common.py b/tests/models/multimodal/generation/test_common.py
index de74acf3b8a5..5504c417fda4 100644
--- a/tests/models/multimodal/generation/test_common.py
+++ b/tests/models/multimodal/generation/test_common.py
@@ -9,11 +9,13 @@ from collections import defaultdict
 from pathlib import PosixPath
 
 import pytest
+from packaging.version import Version
 from transformers import (
     AutoModel,
     AutoModelForImageTextToText,
     AutoModelForTextToWaveform,
 )
+from transformers import __version__ as TRANSFORMERS_VERSION
 
 from vllm.platforms import current_platform
 from vllm.utils.func_utils import identity
@@ -851,6 +853,12 @@ VLM_TEST_SETTINGS = {
                 limit_mm_per_prompt={"image": 4},
             )
         ],
+        marks=[
+            pytest.mark.skipif(
+                Version(TRANSFORMERS_VERSION) == Version("4.57.1"),
+                reason="This model is broken in Transformers v4.57.1",
+            )
+        ],
     ),
     # regression test for https://github.com/vllm-project/vllm/issues/15122
     "qwen2_5_vl-windows-attention": VLMTestInfo(