vLLM/recipes/models/XiaomiMiMo/MiMo-V2-Flash.yaml

meta:
  title: "MiMo-V2-Flash"
  slug: "mimo-v2-flash"
  provider: "MiMo (Xiaomi)"
  description: "Xiaomi's MoE reasoning model (309B total / 15B active) with hybrid attention and MTP for fast agentic workflows"
  date_added: 2025-12-20
  date_updated: 2026-10-08
  tasks:
    - text
  related_recipes: []
  hardware:
    h200: verified
    mi300x: verified
    mi325x: verified
    mi355x: verified

model:
  model_id: "XiaomiMiMo/MiMo-V2-Flash"
  min_vllm_version: "0.20.0"
  architecture: moe
  parameter_count: "309B"
  active_parameters: "15B"
  context_length: 262144
  base_args:
    - "--trust-remote-code"
    - "--generation-config"
    - "vllm"
  base_env: {}

features:
  tool_calling:
    description: "MiMo tool-call parser"
    args:
      - "--tool-call-parser"
      - "mimo"
  reasoning:
    description: "Qwen3 reasoning parser"
    args:
      - "--reasoning-parser"
      - "qwen3"

opt_in_features: []

variants:
  default:
    precision: fp8
    vram_minimum_gb: 371
    description: "Native FP8 weights; 4x H200 recommended with TP4"
    extra_args:
      - "--tensor-parallel-size"
      - "4"
      - "--gpu-memory-utilization"
      - "0.9"

compatible_strategies:
  - single_node_tp
  - single_node_tep
  - multi_node_tp
  - multi_node_tp_pp
  - multi_node_tep
  - multi_node_dep
  - pd_cluster

hardware_overrides:
  amd:
    # AITER is disabled for MiMo-V2-Flash (regression when enabled).
    extra_args: []
    extra_env:
      VLLM_ROCM_USE_AITER: "0"
strategy_overrides: {}

guide: |
  ## Overview

  [MiMo-V2-Flash](https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash) is a MoE language model
  with 309B total parameters and 15B active. Designed for high-speed reasoning and agentic
  workflows, it features hybrid attention and Multi-Token Prediction (MTP) to reduce
  inference cost.

  ## Prerequisites

  - Hardware: 4x H200 (TP4) or equivalent aggregate VRAM (~320 GB with FP8)
  - vLLM >= 0.20.0

  ### Install vLLM (NVIDIA)

  ```bash
  uv venv
  source .venv/bin/activate
  uv pip install vllm --torch-backend auto
  ```

  ### Install vLLM (AMD ROCm MI300X/MI325X/MI355X)

  ```bash
  uv pip install vllm --extra-index-url https://wheels.vllm.ai/rocm/
  ```

  ## Launch commands

  Basic TP4:

  ```bash
  vllm serve XiaomiMiMo/MiMo-V2-Flash \
    --host 0.0.0.0 --port 9001 --seed 1024 \
    --served-model-name mimo_v2_flash \
    --tensor-parallel-size 4 \
    --trust-remote-code \
    --gpu-memory-utilization 0.9 \
    --generation-config vllm
  ```

  With tool calling + reasoning:

  ```bash
  vllm serve XiaomiMiMo/MiMo-V2-Flash \
    --tensor-parallel-size 4 --trust-remote-code --gpu-memory-utilization 0.9 \
    --tool-call-parser mimo \
    --reasoning-parser qwen3 \
    --generation-config vllm \
    --served-model-name mimo_v2_flash
  ```

  DP + TP + EP:

  ```bash
  vllm serve XiaomiMiMo/MiMo-V2-Flash \
    --data-parallel-size 2 \
    --tensor-parallel-size 4 \
    --enable-expert-parallel \
    --trust-remote-code \
    --gpu-memory-utilization 0.9 \
    --generation-config vllm \
    --served-model-name mimo_v2_flash
  ```

  AMD:

  ```bash
  export VLLM_ROCM_USE_AITER=0
  vllm serve XiaomiMiMo/MiMo-V2-Flash --tensor-parallel-size 4 \
    --trust-remote-code --gpu-memory-utilization 0.9 --generation-config vllm
  ```

  Tunable flags:
  - `--max-model-len=65536` works well; max is 128K.
  - `--max-num-batched-tokens=32768` for prompt-heavy; 16K/8K for lower latency.
  - `--gpu-memory-utilization=0.95` to maximize KV cache.

  ## Client Usage

  ```bash
  curl -X POST http://localhost:9001/v1/chat/completions \
    -H "Content-Type: application/json" \
    -d '{
      "model": "mimo_v2_flash",
      "messages": [{"role": "user", "content": "Hello MiMo!"}],
      "chat_template_kwargs": {"enable_thinking": true}
    }'
  ```

  Set `"enable_thinking": false` (or omit the kwargs) to disable thinking mode.

  ## Benchmarking

  ```bash
  vllm bench serve \
    --model XiaomiMiMo/MiMo-V2-Flash \
    --dataset-name random --random-input-len 8000 --random-output-len 1000 \
    --request-rate 3 --num-prompts 1800 --ignore-eos
  ```

  ## Accuracy (GSM8K)

  Reported 5-shot `exact_match`: flexible 0.9128, strict 0.9075.

  ## References

  - [MiMo-V2-Flash on Hugging Face](https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash)