diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py
index 988144197ba..2d4f1959c38 100644
--- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py
+++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py
@@ -817,6 +817,7 @@ def test_trtllm_bench(llm_root):  # noqa: F811
 
 
 @pytest.mark.no_xdist
+@pytest.mark.skip(reason="https://nvbugs/5458798")
 def test_trtllm_bench_backend_comparison(llm_root):  # noqa: F811
     """Test that compares autodeploy backend performance against pytorch backend
     with given relative and absolute thresholds.