diff --git a/python/sglang/srt/layers/quantization/modelopt_quant.py b/python/sglang/srt/layers/quantization/modelopt_quant.py
index f7fc3d026..dc9501133 100755
--- a/python/sglang/srt/layers/quantization/modelopt_quant.py
+++ b/python/sglang/srt/layers/quantization/modelopt_quant.py
@@ -66,12 +66,15 @@ if TYPE_CHECKING:
         StandardDispatchOutput,
     )
 
+fp4_quantize = None
 try:
     if is_sm120_supported():
-        from flashinfer import fp4_quantize
+        try:
+            from flashinfer import fp4_quantize
+        except ImportError:
+            from sgl_kernel import scaled_fp4_quant as fp4_quantize
     else:
         from sgl_kernel import scaled_fp4_quant as fp4_quantize
-
 except ImportError:
     fp4_quantize = None