diff --git a/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py b/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py
index a9b30b780..799854479 100644
--- a/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py
+++ b/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py
@@ -264,6 +264,7 @@ def is_flashinfer_supporting_global_sf(backend: FlashinferMoeBackend | None) ->
     backends_supporting_global_sf = (
         FlashinferMoeBackend.CUTLASS,
         FlashinferMoeBackend.TENSORRT_LLM,
+        FlashinferMoeBackend.CUTEDSL,
     )
     return backend in backends_supporting_global_sf