diff --git a/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/agg.yaml b/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/agg.yaml
index 72d65baae7..f79c92e72e 100644
--- a/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/agg.yaml
+++ b/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/agg.yaml
@@ -48,7 +48,7 @@ roles:
       SGLANG_HICACHE_DEBUG_LOG: '1'
       SGLANG_HICACHE_DEBUG_SAMPLE_RATE: '16384'
       SGLANG_MOE_NVFP4_DISPATCH: '1'
-      SGLANG_NVFP4_CKPT_FP8_NEXTN_MOE: '1'
+      SGLANG_NVFP4_CKPT_FP8_NEXTN_MOE: '0'
       SGLANG_CLIP_MAX_NEW_TOKENS_ESTIMATION: '8'
       PIP_BREAK_SYSTEM_PACKAGES: '1'
       SGLANG_DG_CACHE_DIR: /deepgemm_cache
diff --git a/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/disagg-mtp-variants.yaml b/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/disagg-mtp-variants.yaml
index 36ef2c32ff..9bebf23134 100644
--- a/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/disagg-mtp-variants.yaml
+++ b/inferencex-e2e/benchmarks/multi_node/srt-slurm-recipes/glm5.2/sglang/gb300-fp4/agentx/disagg-mtp-variants.yaml
@@ -87,6 +87,7 @@ base:
         speculative-num-steps: 1
         speculative-eagle-topk: 1
         speculative-num-draft-tokens: 2
+        watchdog-timeout: 1800
         enable-metrics: true
         enable-cache-report: true
     decode:
@@ -106,7 +107,7 @@ base:
         SGLANG_DISABLE_TP_MEMORY_INBALANCE_CHECK: '1'
         SGLANG_DEEPEP_NUM_MAX_DISPATCH_TOKENS_PER_RANK: '512'
         SGLANG_MOE_NVFP4_DISPATCH: '1'
-        SGLANG_NVFP4_CKPT_FP8_NEXTN_MOE: '1'
+        SGLANG_NVFP4_CKPT_FP8_NEXTN_MOE: '0'
         SGLANG_ENABLE_THINKING: '1'
         SGLANG_CLIP_MAX_NEW_TOKENS_ESTIMATION: '8'
         SGLANG_REASONING_EFFORT: max
@@ -147,6 +148,7 @@ base:
         model-loader-extra-config: '{"enable_multithread_load": true}'
         mem-fraction-static: 0.9
         disaggregation-decode-extra-slots: 0
+        watchdog-timeout: 1800
         enable-metrics: true
         enable-cache-report: true
   health_check:
diff --git a/inferencex-e2e/perf-changelog.yaml b/inferencex-e2e/perf-changelog.yaml
index c2430c0b2e..88b9fbf6dd 100644
--- a/inferencex-e2e/perf-changelog.yaml
+++ b/inferencex-e2e/perf-changelog.yaml
@@ -9245,3 +9245,15 @@
     - "Disable FP8 conversion of the GLM-5.2 NextN/MTP draft MoE on B200 AgentX so the draft runs at its shipped precision; images, topology, workload and acceptance are unchanged."
     - "关闭 B200 GLM-5.2 AgentX 对 NextN/MTP draft MoE 的 FP8 转换，使 draft 保持原始发布精度；镜像、拓扑、工作负载和验收不变。"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/3400
+
+- config-keys:
+    - glm5.2-fp4-gb300-dynamo-sglang-agentic-agg
+    - glm5.2-fp4-gb300-dynamo-sglang-agentic-disagg
+  scenario-type:
+    - agentic-coding
+  description:
+    - "Disable FP8 conversion of the GLM-5.2 NextN/MTP draft MoE on GB300 AgentX so the draft runs at its shipped precision; images, topology, workload and acceptance are unchanged."
+    - "Disaggregated prefill and decode use the 1800 s scheduler watchdog of the GLM-5.2 B200/GB200 recipes instead of the 300 s default."
+    - "关闭 GB300 GLM-5.2 AgentX 对 NextN/MTP draft MoE 的 FP8 转换，使 draft 保持原始发布精度；镜像、拓扑、工作负载和验收不变。"
+    - "分离式 prefill 和 decode 采用 GLM-5.2 B200/GB200 配方的 1800 秒调度 watchdog，替代默认 300 秒。"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/3402