diff --git a/vllm/model_executor/layers/quantization/inc.py b/vllm/model_executor/layers/quantization/inc.py
index 4457555c0..62794613d 100644
--- a/vllm/model_executor/layers/quantization/inc.py
+++ b/vllm/model_executor/layers/quantization/inc.py
@@ -413,6 +413,8 @@ class INCConfig(QuantizationConfig):
         return None
 
     def apply_xpu_w4a16_quant_layer(self, layer, prefix: str):
+        from vllm.model_executor.layers.fused_moe import FusedMoE
+
         weight_bits, group_size, sym = self.get_layer_config(layer, prefix)
 
         if not self.check_quantized(weight_bits):
@@ -430,6 +432,19 @@ class INCConfig(QuantizationConfig):
             raise NotImplementedError(
                 "INC W4A16 on XPU only supports symmetric quantization for now."
             )
+        if isinstance(layer, FusedMoE):
+            from vllm.model_executor.layers.quantization.moe_wna16 import (
+                MoeWNA16Config,
+            )
+
+            config = {
+                "quant_method": "gptq",
+                "bits": weight_bits,
+                "group_size": group_size,
+                "sym": sym,
+                "lm_head": False,
+            }
+            return MoeWNA16Config.from_config(config).get_quant_method(layer, prefix)
         if isinstance(layer, (LinearBase, ParallelLMHead)):
             return INCXPULinearMethod(
                 weight_bits=weight_bits,