diff --git a/benchmarks/single_node/agentic/minimaxm3_fp4_b200_mtp.sh b/benchmarks/single_node/agentic/minimaxm3_fp4_b200_mtp.sh
index be705c0c2..23a2c1010 100755
--- a/benchmarks/single_node/agentic/minimaxm3_fp4_b200_mtp.sh
+++ b/benchmarks/single_node/agentic/minimaxm3_fp4_b200_mtp.sh
@@ -194,6 +194,7 @@ VLLM_CMD=(
     --max-cudagraph-capture-size 512
     --max-num-batched-tokens 16384
     --stream-interval 20
+    --long-prefill-token-threshold 512 \
     --trust-remote-code
     --speculative-config "$SPEC_CONFIG"
     "${OFFLOAD_ARGS[@]}"
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index 90953bc2d..578a2c523 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -5688,3 +5688,9 @@
     - "Use the official lmsysorg/sglang:v0.5.16-rocm720-mi30x image, enable SGLang prompt-cache reporting, and pass the backend Prometheus metrics endpoint explicitly to AIPerf"
     - "Allow up to 30 minutes for healthy AgentX responses admitted near the end of the measurement window to drain before AIPerf finalizes profile metric coverage"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2528
+
+- config-keys:
+    - minimaxm3-fp4-b200-vllm-agentic-mtp
+  description:
+    - "Set --long-prefill-token-threshold 512 for MiniMax M3 NVFP4 b200 AgentX MTP"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2538