diff --git a/olive/passes/onnx/quantization.py b/olive/passes/onnx/quantization.py
index 42504b90b..9607b8054 100644
--- a/olive/passes/onnx/quantization.py
+++ b/olive/passes/onnx/quantization.py
@@ -608,7 +608,6 @@ def _default_config(cls, accelerator_spec: AcceleratorSpec) -> Dict[str, PassCon
             config["activation_type"].search_defaults = Categorical(["QInt8", "QUInt8", "QUInt16", "QInt16"])
             config["weight_type"].search_defaults = Categorical(["QInt8", "QUInt8", "QUInt16", "QInt16"])
             config["prepare_qnn_config"].default_value = True
-            config["quant_preprocess"].default_value = False
             # in QNN EP, the default value WeightSymmetric is None
             # but in base quantizer, the default value is True.
             config["WeightSymmetric"].default_value = None