[Bugfix] Allow flashinfer_cutlass as a clamped NVFP4 MoE backend (#46492)

Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com> Signed-off-by: Michael Goin <mgoin64@gmail.com> Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: Michael Goin <mgoin64@gmail.com>

Gabriel Wu committed Jun 24, 2026 at 03:43 UTC 0d4d164488cb29bdc9fbfe3cd943da4b854ae19c
1 file changed +1
vllm/model_executor/layers/fused_moe/oracle/nvfp4.py
+1
@@ -174,6 +174,7 @@ def select_nvfp4_moe_backend(
174
175 NVFP4_BACKENDS_WITH_CLAMP = {
176 NvFp4MoeBackend.FLASHINFER_TRTLLM,
177 + NvFp4MoeBackend.FLASHINFER_CUTLASS,
178 NvFp4MoeBackend.MARLIN,
179 }
180