diff --git a/benchmarks/single_node/fixed_seq_len/kimik2.5_fp4_b200.sh b/benchmarks/single_node/fixed_seq_len/kimik2.5_fp4_b200.sh
index 6c413137a..793bf2553 100644
--- a/benchmarks/single_node/fixed_seq_len/kimik2.5_fp4_b200.sh
+++ b/benchmarks/single_node/fixed_seq_len/kimik2.5_fp4_b200.sh
@@ -12,6 +12,20 @@ check_env_vars \
     RANDOM_RANGE_RATIO \
     RESULT_FILENAME
 
+PARALLEL_ARGS=(--tensor-parallel-size "$TP" --data-parallel-size 1)
+GMU=0.90
+PREFILL_SCHEDULE_ARGS=()
+if [ "${DP_ATTENTION:-false}" = "true" ]; then
+    PARALLEL_ARGS=(--tensor-parallel-size 1 --data-parallel-size "$TP")
+    GMU=0.80
+    PREFILL_SCHEDULE_ARGS=(--prefill-schedule-interval 4)
+fi
+
+EP_ARGS=()
+if [ "${EP_SIZE:-1}" -gt 1 ]; then
+    EP_ARGS=(--enable-expert-parallel)
+fi
+
 if [[ -n "$SLURM_JOB_ID" ]]; then
   echo "JOB $SLURM_JOB_ID running on $SLURMD_NODENAME"
 fi
@@ -22,6 +36,8 @@ nvidia-smi
 
 export TORCH_CUDA_ARCH_LIST="10.0"
 export PYTHONNOUSERSITE=1
+export VLLM_USE_V2_MODEL_RUNNER=0
+export VLLM_FLASHINFER_AUTOTUNE_SKIP_OPS=""
 
 SERVER_LOG=/workspace/server.log
 
@@ -41,17 +57,21 @@ export VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0
 
 set -x
 vllm serve $MODEL --host 0.0.0.0 --port $PORT \
---tensor-parallel-size=$TP \
---gpu-memory-utilization 0.90 \
+"${PARALLEL_ARGS[@]}" \
+"${EP_ARGS[@]}" \
+"${PREFILL_SCHEDULE_ARGS[@]}" \
+--gpu-memory-utilization "$GMU" \
 --max-model-len $MAX_MODEL_LEN \
 --max-num-seqs $CONC \
 --reasoning-parser kimi_k2 \
 --tool-call-parser kimi_k2 \
 --compilation_config.pass_config.fuse_allreduce_rms true \
 --kv-cache-dtype fp8 \
---max-cudagraph-capture-size 2048 \
---max-num-batched-tokens "$((ISL * 2 ))" \
---stream-interval 20 --no-enable-prefix-caching \
+--max-cudagraph-capture-size "$((CONC * 2))" \
+--stream-interval 32 \
+--attention-config '{"mla_prefill_backend":"FLASHINFER","use_prefill_query_quantization":true}' \
+--linear-backend flashinfer_cutlass \
+--no-enable-prefix-caching \
 --trust-remote-code > $SERVER_LOG 2>&1 &
 
 SERVER_PID=$!
diff --git a/configs/nvidia-master.yaml b/configs/nvidia-master.yaml
index 51a650397..ac60822d5 100644
--- a/configs/nvidia-master.yaml
+++ b/configs/nvidia-master.yaml
@@ -1416,7 +1416,7 @@ kimik2.5-int4-h200-vllm:
 # Kimi-K2.5 FP4 B200 vLLM recipe as-is until B300-specific tuning is available.
 
 kimik2.5-fp4-b200-vllm:
-  image: vllm/vllm-openai:v0.22.0
+  image: vllm/vllm-openai:nightly-e2fa28594f7baad142a426b0b6a2cfe2c79201c7
   model: nvidia/Kimi-K2.5-NVFP4
   model-prefix: kimik2.5
   runner: b200
@@ -1428,8 +1428,31 @@ kimik2.5-fp4-b200-vllm:
     - isl: 8192
       osl: 1024
       search-space:
-      - { tp: 8, ep: 1, conc-start: 1, conc-end: 4 }
-      - { tp: 4, ep: 1, conc-start: 1, conc-end: 128 }
+      - { tp: 8, ep: 1, conc-start: 1, conc-end: 512 }
+      - { tp: 4, ep: 1, conc-start: 1, conc-end: 512 }
+      - { tp: 8, ep: 8, dp-attn: false, conc-start: 1, conc-end: 512 }
+      - { tp: 4, ep: 4, dp-attn: false, conc-start: 1, conc-end: 512 }
+      - { tp: 8, ep: 8, dp-attn: true, conc-list: [512, 1024] }
+
+# NOTE: At the time of submission, https://docs.vllm.ai/projects/recipes/en/latest/moonshotai/Kimi-K2.5.html
+# does not have a B300-specific recipe, so this config reuses the existing
+# Kimi-K2.5 FP4 B200 vLLM recipe as-is until B300-specific tuning is available.
+m:
+  image: vllm/vllm-openai:nightly-e2fa28594f7baad142a426b0b6a2cfe2c79201c7
+  model: nvidia/Kimi-K2.5-NVFP4
+  model-prefix: kimik2.5
+  runner: b200
+  precision: fp4
+  framework: vllm
+  multinode: false
+  scenarios:
+    fixed-seq-len:
+    - isl: 8192
+      osl: 1024
+      search-space:
+      - { tp: 8, ep: 1, conc-start: 1, conc-end: 512 }
+      - { tp: 4, ep: 1, conc-start: 1, conc-end: 512 }
+      - { tp: 8, ep: 8, dp-attn: true, conc-list: [512, 1024] }
 
 # NOTE: At the time of submission, https://docs.vllm.ai/projects/recipes/en/latest/moonshotai/Kimi-K2.5.html
 # does not have a B300-specific recipe, so this config reuses the existing
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index c855c972c..a892d7e6d 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -5449,3 +5449,9 @@
     - "Enable prefill-only INT4 quick-reduce: set VLLM_ROCM_QUICK_REDUCE_QUANTIZATION=INT4 and VLLM_ROCM_QUICK_REDUCE_MAX_SIZE_BYTES_MB=2048 on the prefill workers via a new prefill_env channel (mirrors the existing decode_env path in server_vllm.sh)."
     - "Cap the 1P1D TP4 concurrency sweep at 256 (was 512); drop the 2P1D TP4 layout (128/256/512) as it is CI-flaky with negligible curve impact."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1943
+
+- config-keys:
+    - kimik2.5-fp4-b200-vllm
+  description:
+    - "Kimi K2.5 NVFP4 B200 vLLM: nightly image, extend sweep space with TP/DEP arms, DEP gmu 0.80"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2443