diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-dep8-eagle3-c64-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-dep8-eagle3-c64-8k1k.yaml
new file mode 100644
index 0000000000..72fcd6cc2e
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-dep8-eagle3-c64-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-dep8-mxfp8-8k1k-eagle3-c64"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 2
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "64"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c1-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c1-8k1k.yaml
new file mode 100644
index 0000000000..990e540075
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c1-8k1k.yaml
@@ -0,0 +1,113 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-tp4-mxfp8-8k1k-eagle3-c1"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 1
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 4
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 4
+      enable-expert-parallel: false
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "1"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c8-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c8-8k1k.yaml
new file mode 100644
index 0000000000..ff8f84cf47
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c8-8k1k.yaml
@@ -0,0 +1,113 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-tp4-mxfp8-8k1k-eagle3-c8"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 1
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 4
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 4
+      enable-expert-parallel: false
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "8"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c1-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c1-8k1k.yaml
new file mode 100644
index 0000000000..4fa642e3d9
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c1-8k1k.yaml
@@ -0,0 +1,113 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-tp8-mxfp8-8k1k-eagle3-c1"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 2
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 8
+      enable-expert-parallel: false
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "1"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c4-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c4-8k1k.yaml
new file mode 100644
index 0000000000..280819ae5a
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c4-8k1k.yaml
@@ -0,0 +1,113 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-tp8-mxfp8-8k1k-eagle3-c4"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 2
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 8
+      enable-expert-parallel: false
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "4"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c8-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c8-8k1k.yaml
new file mode 100644
index 0000000000..94322297f0
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c8-8k1k.yaml
@@ -0,0 +1,113 @@
+name: "minimax-m3-vllm-disagg-gb300-1p1d-dep2-tp8-mxfp8-8k1k-eagle3-c8"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 2
+  prefill_workers: 1
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 8
+      enable-expert-parallel: false
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "8"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/2p1d-dep2-dep8-eagle3-c512-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/2p1d-dep2-dep8-eagle3-c512-8k1k.yaml
new file mode 100644
index 0000000000..e75a8eb06a
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/2p1d-dep2-dep8-eagle3-c512-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-2p1d-dep2-dep8-mxfp8-8k1k-eagle3-c512"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 1
+  decode_nodes: 2
+  prefill_workers: 2
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "512"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/3p1d-dep2-dep8-eagle3-c256-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/3p1d-dep2-dep8-eagle3-c256-8k1k.yaml
new file mode 100644
index 0000000000..3a7753467c
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/3p1d-dep2-dep8-eagle3-c256-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-3p1d-dep2-dep8-mxfp8-8k1k-eagle3-c256"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 2
+  decode_nodes: 2
+  prefill_workers: 3
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "256"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c1024-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c1024-8k1k.yaml
new file mode 100644
index 0000000000..4b25736627
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c1024-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-4p1d-dep2-dep8-mxfp8-8k1k-eagle3-c1024"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 2
+  decode_nodes: 2
+  prefill_workers: 4
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "1024"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c2048-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c2048-8k1k.yaml
new file mode 100644
index 0000000000..c9537a9fcd
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c2048-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-4p1d-dep2-dep8-mxfp8-8k1k-eagle3-c2048"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 2
+  decode_nodes: 2
+  prefill_workers: 4
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "2048"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/6p1d-dep2-dep8-eagle3-c2048-8k1k.yaml b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/6p1d-dep2-dep8-eagle3-c2048-8k1k.yaml
new file mode 100644
index 0000000000..601864149f
--- /dev/null
+++ b/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/6p1d-dep2-dep8-eagle3-c2048-8k1k.yaml
@@ -0,0 +1,116 @@
+name: "minimax-m3-vllm-disagg-gb300-6p1d-dep2-dep8-mxfp8-8k1k-eagle3-c2048"
+
+model:
+  path: "minimax-m3-mxfp8"
+  container: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  precision: "fp8"
+
+identity:
+  model:
+    repo: "MiniMaxAI/MiniMax-M3-MXFP8"
+    revision: "c5454eb03678d8710e54a4e0fc681b9f3b4a3dba"
+  container:
+    image: "vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7"
+  frameworks:
+    dynamo: "1.4.0.dev20260730"
+    vllm: "0.26.1rc1.dev255+g5e35a6f4f"
+
+dynamo:
+  install: true
+  version: "1.4.0.dev20260730"
+  request_plane: "nats"
+
+health_check:
+  max_attempts: 720
+  interval_seconds: 10
+
+sbatch_directives:
+  mem: "0"
+  cpus-per-task: "72"
+
+srun_options:
+  mem: "0"
+
+resources:
+  gpu_type: "gb300"
+  gpus_per_node: 4
+  prefill_nodes: 3
+  decode_nodes: 2
+  prefill_workers: 6
+  decode_workers: 1
+  gpus_per_prefill: 2
+  gpus_per_decode: 8
+
+frontend:
+  type: "dynamo"
+  enable_multiple_frontends: false
+
+backend:
+  type: "vllm"
+  connector: null
+
+  prefill_environment: &worker-environment
+    VLLM_ENGINE_READY_TIMEOUT_S: "3600"
+    VLLM_FLOAT32_MATMUL_PRECISION: "high"
+    VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl"
+    UCX_CUDA_IPC_ENABLE_MNNVL: "y"
+    UCX_MODULE_DIR: "/usr/local/lib/python3.12/dist-packages/nixl_cu13.libs/ucx"
+    UCX_RNDV_PIPELINE_ERROR_HANDLING: "y"
+    NCCL_CUMEM_ENABLE: "1"
+    NCCL_MNNVL_ENABLE: "1"
+    NCCL_NVLS_ENABLE: "1"
+
+  decode_environment: *worker-environment
+
+  vllm_config:
+    prefill:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 2
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-cudagraph-capture-size: 2048
+      max-num-batched-tokens: 16384
+
+    decode:
+      no-enable-flashinfer-autotune: true
+      kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}'
+      attention-config: '{"backend": "FLASHINFER", "use_trtllm_attention": true, "indexer_kv_dtype": "fp8", "minimax_m3_msa_decode_backend": "cutlass"}'
+      speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
+      tensor-parallel-size: 1
+      data-parallel-size: 8
+      data-parallel-rpc-port: 13345
+      enable-expert-parallel: true
+      trust-remote-code: true
+      no-enable-prefix-caching: true
+      block-size: 128
+      gpu-memory-utilization: 0.90
+      max-model-len: 9472
+      language-model-only: true
+      kv-cache-dtype: "fp8"
+      stream-interval: 32
+      max-num-seqs: 1024
+      max-num-batched-tokens: 16384
+      max-cudagraph-capture-size: 2048
+      compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY"}'
+
+benchmark:
+  type: "sa-bench"
+  isl: 8192
+  osl: 1024
+  concurrencies: "2048"
+  req_rate: "inf"
+  num_warmup_mult: 2
+  random_range_ratio: 0.8
+  use_chat_template: true
diff --git a/configs/nvidia-master.yaml b/configs/nvidia-master.yaml
index 39087de3ae..d6ea947df5 100644
--- a/configs/nvidia-master.yaml
+++ b/configs/nvidia-master.yaml
@@ -8008,6 +8008,178 @@ minimaxm3-fp4-b300-dynamo-vllm-mtp-legacy-dep4:
 # MiniMax-M3 GB300 disagg sweep — refreshed recipe set (no Marlin variants).
 # All prefill DEP2 (TP1 DP2 EP, 2 GPU/worker). Decode: DEP4, TEP8, DEP8, TEP4.
 # 4 GPU/node (GB300 NVL72). kv-cache-dtype=fp8. srun_options mem=0 required.
+minimaxm3-fp8-gb300-dynamo-vllm-mtp:
+  image: vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7
+  model: MiniMaxAI/MiniMax-M3-MXFP8
+  model-prefix: minimaxm3
+  runner: gb300-nv
+  precision: fp8
+  framework: dynamo-vllm
+  router: { name: dynamo-router, version: "1.4.0.dev20260730" }
+  kv-p2p-transfer: nixl
+  multinode: true
+  disagg: true
+  scenarios:
+    fixed-seq-len:
+    - isl: 8192
+      osl: 1024
+      search-space:
+      - spec-decoding: "mtp"
+        conc-list: [1]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c1-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 4
+          ep: 1
+          dp-attn: false
+      - spec-decoding: "mtp"
+        conc-list: [8]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp4-eagle3-c8-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 4
+          ep: 1
+          dp-attn: false
+      - spec-decoding: "mtp"
+        conc-list: [1]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c1-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+      - spec-decoding: "mtp"
+        conc-list: [4]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c4-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+      - spec-decoding: "mtp"
+        conc-list: [8]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-tp8-eagle3-c8-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+      - spec-decoding: "mtp"
+        conc-list: [64]
+        prefill:
+          num-worker: 1
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/1p1d-dep2-dep8-eagle3-c64-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+      - spec-decoding: "mtp"
+        conc-list: [512]
+        prefill:
+          num-worker: 2
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/2p1d-dep2-dep8-eagle3-c512-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+      - spec-decoding: "mtp"
+        conc-list: [256]
+        prefill:
+          num-worker: 3
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/3p1d-dep2-dep8-eagle3-c256-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+      - spec-decoding: "mtp"
+        conc-list: [1024]
+        prefill:
+          num-worker: 4
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c1024-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+      - spec-decoding: "mtp"
+        conc-list: [2048]
+        prefill:
+          num-worker: 4
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/4p1d-dep2-dep8-eagle3-c2048-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+      - spec-decoding: "mtp"
+        conc-list: [2048]
+        prefill:
+          num-worker: 6
+          tp: 2
+          ep: 2
+          dp-attn: true
+          additional-settings:
+          - "CONFIG_FILE=recipes/vllm/minimax-m3-gb300-fp8/8k1k/mtp/6p1d-dep2-dep8-eagle3-c2048-8k1k.yaml"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 8
+          dp-attn: true
+
+
 minimaxm3-fp8-gb300-dynamo-vllm:
   image: vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9
   model: MiniMaxAI/MiniMax-M3-MXFP8
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index 43dc7fdb6f..41bd57494c 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -5431,3 +5431,12 @@
   description:
     - "Add MiniMax M3 NVFP4 B200 disagg EAGLE3 MTP recipes, mirrored from B300"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2462
+
+- config-keys:
+    - minimaxm3-fp8-gb300-dynamo-vllm-mtp
+  scenario-type:
+    - fixed-seq-len
+  description:
+    - "Add MiniMax-M3-MXFP8 GB300 8k1k disaggregated Dynamo-vLLM EAGLE3 MTP Pareto on vLLM nightly 5e35a6f4f"
+    - "The six dep8 decode configurations enable the CUTLASS MSA sparse decode backend and FULL_DECODE_ONLY cudagraph mode; the five low-concurrency c1/c4/c8 configurations use neither"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2486
diff --git a/runners/launch_gb300-nv.sh b/runners/launch_gb300-nv.sh
index 7232174d92..36fcad9e50 100644
--- a/runners/launch_gb300-nv.sh
+++ b/runners/launch_gb300-nv.sh
@@ -226,7 +226,11 @@ elif [[ $FRAMEWORK == "dynamo-sglang" && $MODEL_PREFIX == "qwen3.5" ]]; then
 elif [[ $FRAMEWORK == "dynamo-vllm" && $MODEL_PREFIX == "minimaxm3" ]]; then
     git clone https://github.com/NVIDIA/srt-slurm.git "$SRT_REPO_DIR"
     cd "$SRT_REPO_DIR"
-    git checkout sa-submission-q2-2026
+    if [[ "${SPEC_DECODING:-}" == "mtp" ]]; then
+        git checkout v1.0.38
+    else
+        git checkout sa-submission-q2-2026
+    fi
     mkdir -p recipes/vllm/minimax-m3-gb300-fp8
     cp -rT "$GITHUB_WORKSPACE/benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m3-gb300-fp8" recipes/vllm/minimax-m3-gb300-fp8
 elif [[ $FRAMEWORK == "dynamo-vllm" && $MODEL_PREFIX == "kimik2.5" && $PRECISION == "fp4" ]]; then