diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p1d-dep2-tp4-eagle3-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p1d-dep2-tp4-eagle3-8k1k.yaml
index 0890ace22f..165301c53c 100644
--- a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p1d-dep2-tp4-eagle3-8k1k.yaml
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p1d-dep2-tp4-eagle3-8k1k.yaml
@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-b300-1p1d-dep2-tp4-fp4-8k1k-eagle3"
 
 model:
   path: "nvidia/MiniMax-M3-NVFP4"
-  container: "vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2"
+  container: "vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
   precision: fp4
 
 resources:
@@ -31,12 +31,10 @@ backend:
 
   prefill_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   decode_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   vllm_config:
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p2d-dep2-tp4-eagle3-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p2d-dep2-tp4-eagle3-8k1k.yaml
index 19cdeb0793..b32295e325 100644
--- a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p2d-dep2-tp4-eagle3-8k1k.yaml
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p2d-dep2-tp4-eagle3-8k1k.yaml
@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-b300-1p2d-dep2-tp4-fp4-8k1k-eagle3"
 
 model:
   path: "nvidia/MiniMax-M3-NVFP4"
-  container: "vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2"
+  container: "vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
   precision: fp4
 
 resources:
@@ -31,12 +31,10 @@ backend:
 
   prefill_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   decode_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   vllm_config:
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p4d-dep2-tp4-eagle3-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p4d-dep2-tp4-eagle3-8k1k.yaml
index 858576055b..ff1b0c73c7 100644
--- a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p4d-dep2-tp4-eagle3-8k1k.yaml
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p4d-dep2-tp4-eagle3-8k1k.yaml
@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-b300-1p4d-dep2-tp4-fp4-8k1k-eagle3"
 
 model:
   path: "nvidia/MiniMax-M3-NVFP4"
-  container: "vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2"
+  container: "vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
   precision: fp4
 
 resources:
@@ -31,12 +31,10 @@ backend:
 
   prefill_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   decode_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   vllm_config:
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p6d-dep2-tp4-eagle3-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p6d-dep2-tp4-eagle3-8k1k.yaml
index b4da6d26e4..c5f096d64d 100644
--- a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p6d-dep2-tp4-eagle3-8k1k.yaml
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/1p6d-dep2-tp4-eagle3-8k1k.yaml
@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-b300-1p6d-dep2-tp4-fp4-8k1k-eagle3"
 
 model:
   path: "nvidia/MiniMax-M3-NVFP4"
-  container: "vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2"
+  container: "vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
   precision: fp4
 
 resources:
@@ -31,12 +31,10 @@ backend:
 
   prefill_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   decode_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   vllm_config:
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/2p3d-dep2-tp4-eagle3-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/2p3d-dep2-tp4-eagle3-8k1k.yaml
index 73aa7c161c..790e6759a4 100644
--- a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/2p3d-dep2-tp4-eagle3-8k1k.yaml
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3/b300-fp4/8k1k/mtp/2p3d-dep2-tp4-eagle3-8k1k.yaml
@@ -2,7 +2,7 @@ name: "minimax-m3-vllm-disagg-b300-2p3d-dep2-tp4-fp4-8k1k-eagle3"
 
 model:
   path: "nvidia/MiniMax-M3-NVFP4"
-  container: "vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2"
+  container: "vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
   precision: fp4
 
 resources:
@@ -31,12 +31,10 @@ backend:
 
   prefill_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   decode_environment:
     VLLM_FLOAT32_MATMUL_PRECISION: high
-    VLLM_FLASHINFER_ALLREDUCE_BACKEND: trtllm
     UCX_TLS: cuda_copy,cuda_ipc,rc
 
   vllm_config:
diff --git a/configs/nvidia-master.yaml b/configs/nvidia-master.yaml
index f70c76ee53..6e4fc8d341 100644
--- a/configs/nvidia-master.yaml
+++ b/configs/nvidia-master.yaml
@@ -6917,7 +6917,7 @@ minimaxm3-fp4-b300-dynamo-vllm-8k1k-legacy-max-tput:
           dp-attn: false
 
 minimaxm3-fp4-b300-dynamo-vllm-mtp:
-  image: vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2
+  image: vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9
   model: nvidia/MiniMax-M3-NVFP4
   model-prefix: minimaxm3
   runner: b300
@@ -7003,6 +7003,23 @@ minimaxm3-fp4-b300-dynamo-vllm-mtp:
           tp: 4
           ep: 1
           dp-attn: false
+
+minimaxm3-fp4-b300-dynamo-vllm-mtp-legacy-dep4:
+  image: vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2
+  model: nvidia/MiniMax-M3-NVFP4
+  model-prefix: minimaxm3
+  runner: b300
+  precision: fp4
+  framework: dynamo-vllm
+  router: { name: dynamo-router, version: "1.3.0.dev20260710" }
+  kv-p2p-transfer: nixl
+  multinode: true
+  disagg: true
+  scenarios:
+    fixed-seq-len:
+    - isl: 8192
+      osl: 1024
+      search-space:
       - spec-decoding: "mtp"
         conc-list: [256, 512]
         prefill:
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index a8aaaa28e0..c3602ff749 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -5061,6 +5061,16 @@
     - "Exclude known-bad nodes mia1-p01-g09,g14 from the disagg node pool"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2301
 
+- config-keys:
+    - minimaxm3-fp4-b300-dynamo-vllm-mtp
+    - minimaxm3-fp4-b300-dynamo-vllm-mtp-legacy-dep4
+  description:
+    - "Refresh five MiniMax-M3 NVFP4 B300 8k1k disaggregated Dynamo-vLLM EAGLE3 recipes from vllm/vllm-openai:nightly-8e981630c9336233ca9de91452f68918bddbc4e2 to vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9"
+    - "Remove the deprecated VLLM_FLASHINFER_ALLREDUCE_BACKEND=trtllm override from the five refreshed prefill and decode environments, matching the image compatibility adjustment established by PR #2120"
+    - "Retain the 2P1D DEP4 recipe on its legacy image and all-reduce settings under a separate matrix key"
+    - "Preserve all six existing 8k1k EAGLE3 topologies, speculative-token settings, chat template, random range ratio, and 12 total concurrency points"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2314
+
 - config-keys:
     - qwen3.5-fp8-gb200-dynamo-sglang-mtp
   description: