diff --git a/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/Dockerfile.pr_validation b/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/Dockerfile.pr_validation
index f4efc63..18608da 100644
--- a/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/Dockerfile.pr_validation
+++ b/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/Dockerfile.pr_validation
@@ -8,15 +8,14 @@ ENV FLASHINFER_DISABLE_VERSION_CHECK=1 \
     PYTHONPATH=/sgl-workspace/sglang/python \
     PYTHONUNBUFFERED=1
 
-COPY sglang_kernel-0.4.6.post1-cp310-abi3-manylinux2014_x86_64.whl /tmp/
 COPY flashinfer_python-0.6.18-py3-none-any.whl /tmp/
 RUN python3 -m pip install --no-deps --force-reinstall \
-      /tmp/sglang_kernel-0.4.6.post1-cp310-abi3-manylinux2014_x86_64.whl \
       /tmp/flashinfer_python-0.6.18-py3-none-any.whl && \
     rm -f /tmp/*.whl
 
-# Replace the image's older Python package with the exact Draft tree. The
-# matching 0.4.6.post1 sglang-kernel wheel is installed above.
+# Replace the image's older Python package with the exact Draft tree. Keep the
+# base image's ABI-matched torch and sglang-kernel binaries: the Draft changes
+# only the Python Kimi/FlashInfer integration and does not require a new kernel.
 RUN rm -rf /sgl-workspace/sglang/python/sglang
 COPY sglang/ /sgl-workspace/sglang/python/sglang/
 COPY test_mxfp4_sm120_cutlass.py /opt/pr-tests/test_mxfp4_sm120_cutlass.py
diff --git a/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/run_pr_representative_point.sh b/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/run_pr_representative_point.sh
index 32cc896..a723fc3 100755
--- a/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/run_pr_representative_point.sh
+++ b/experiments/pro6000/kimi3_pro6000_sglang_sm120_flashinfer_mxfp4/run_pr_representative_point.sh
@@ -19,7 +19,6 @@ SGLANG_SOURCE="${SGLANG_SOURCE:-/data/hzy/src/sglang-kimi-sm120-draft-wt}"
 FLASHINFER_SOURCE="${FLASHINFER_SOURCE:-/data/hzy/src/flashinfer-pr4460-kimi-test-wt}"
 ARTIFACT_DIR="${ARTIFACT_DIR:-/data/hzy/artifacts/sglang-pr-kimi-sm120-300c87a}"
 FLASHINFER_WHEEL="${FLASHINFER_WHEEL:-${ARTIFACT_DIR}/flashinfer_python-0.6.18-py3-none-any.whl}"
-SGLANG_KERNEL_WHEEL="${SGLANG_KERNEL_WHEEL:-${ARTIFACT_DIR}/sglang_kernel-0.4.6.post1-cp310-abi3-manylinux2014_x86_64.whl}"
 BUILD_CONTEXT="${BUILD_CONTEXT:-/tmp/kimi3_sm120_pr_validation_context}"
 FLASHINFER_CACHE="${FLASHINFER_CACHE:-/data/hzy/cache/flashinfer-pr-b525c51}"
 
@@ -79,7 +78,6 @@ check_inputs() {
   local path
   for path in "${MODEL_PATH}" "${SGLANG_SOURCE}/python/sglang" \
     "${FLASHINFER_SOURCE}" "${FLASHINFER_WHEEL}" \
-    "${SGLANG_KERNEL_WHEEL}" \
     "${SCRIPT_DIR}/Dockerfile.pr_validation"; do
     [[ -e "$path" ]] || { echo "ERROR: missing ${path}" >&2; exit 2; }
   done
@@ -103,7 +101,6 @@ prepare_build_context() {
   mkdir -p "${BUILD_CONTEXT}/sglang"
   cp "${SCRIPT_DIR}/Dockerfile.pr_validation" "${BUILD_CONTEXT}/Dockerfile"
   cp "${FLASHINFER_WHEEL}" "${BUILD_CONTEXT}/"
-  cp "${SGLANG_KERNEL_WHEEL}" "${BUILD_CONTEXT}/"
   cp "${SGLANG_SOURCE}/test/registered/unit/layers/quantization/test_mxfp4_sm120_cutlass.py" \
     "${BUILD_CONTEXT}/test_mxfp4_sm120_cutlass.py"
   tar -C "${SGLANG_SOURCE}/python/sglang" -cf - . | \
@@ -111,7 +108,7 @@ prepare_build_context() {
   {
     printf 'sglang=%s\n' "${SGLANG_COMMIT}"
     printf 'flashinfer_implementation=%s\n' "${FLASHINFER_COMMIT}"
-    sha256sum "${FLASHINFER_WHEEL}" "${SGLANG_KERNEL_WHEEL}"
+    sha256sum "${FLASHINFER_WHEEL}"
   } >"${BUILD_CONTEXT}/source_identity.txt"
   cp "${BUILD_CONTEXT}/source_identity.txt" "${RESULT_ROOT}/build/"
   du -sh "${BUILD_CONTEXT}" | tee "${RESULT_ROOT}/build/context_size.txt"