diff --git a/.github/configs/nvidia-master.yaml b/.github/configs/nvidia-master.yaml
index bd204ab36..34a422f1c 100644
--- a/.github/configs/nvidia-master.yaml
+++ b/.github/configs/nvidia-master.yaml
@@ -2632,7 +2632,7 @@ kimik2.5-int4-h200-vllm-agentic:
       - { tp: 8, offloading: cpu,  conc-list: [6, 7, 8, 9, 10, 11, 12, 13, 14] }
 
 kimik2.5-fp4-b200-vllm:
-  image: vllm/vllm-openai:v0.17.0
+  image: vllm/vllm-openai:v0.20.2
   model: nvidia/Kimi-K2.5-NVFP4
   model-prefix: kimik2.5
   runner: b200
diff --git a/benchmarks/single_node/kimik2.5_fp4_b200.sh b/benchmarks/single_node/kimik2.5_fp4_b200.sh
index 9c82d2865..b4e85c14f 100644
--- a/benchmarks/single_node/kimik2.5_fp4_b200.sh
+++ b/benchmarks/single_node/kimik2.5_fp4_b200.sh
@@ -33,6 +33,13 @@ fi
 # Start GPU monitoring (power, temperature, clocks every second)
 start_gpu_monitor
 
+# vLLM v0.20.2+'s CUDA-graph memory profiler pre-reserves ~57 GB/GPU upfront
+# (~32% of total), which collides with --gpu-memory-utilization=0.90 and
+# leaves negative space for the KV cache. Disable the profiler — our 0.90
+# already leaves ~18 GB/GPU as safety net (same pattern as
+# benchmarks/single_node/agentic/kimik2.5_fp4_b200.sh).
+export VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0
+
 set -x
 vllm serve $MODEL --host 0.0.0.0 --port $PORT \
 --tensor-parallel-size=$TP \
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index d053e65c5..a28c66b85 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -2704,3 +2704,10 @@
   description:
     - "Update vLLM image from v0.19.0-cu130 (25d old) to v0.21.0"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1448
+
+- config-keys:
+    - kimik2.5-fp4-b200-vllm
+  description:
+    - "Update vLLM image from v0.20.2 to v0.21.0"
+    - "Add VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0 to disable aggressive CUDA-graph memory profiler that OOMs the KV cache"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1395