huggingface--transformers/docs/source/en/model_doc/mimo_v2_flash.md

This model was contributed to Hugging Face Transformers on 2026-06-30.

MiMo-V2-Flash

Overview

MiMo-V2-Flash is a Mixture-of-Experts (MoE) language model developed by the Xiaomi MiMo team. Designed to establish a new balance between long-context modeling capabilities and inference efficiency, the model is built for strong performance in complex reasoning and agentic tasks. Trained on 27T tokens with native 32k sequence lengths, MiMo-V2-Flash seamlessly supports an extended 256K context window while significantly reducing KV-cache storage compared to standard global attention models.

Key Features

• Hybrid Attention Architecture: Interleaves Sliding Window Attention (SWA) and Global Attention (GA) at a 5:1 ratio, using an aggressive 128-token window. This approach reduces KV-cache storage by nearly 6x while utilizing a learnable attention sink bias to preserve excellent performance on long contexts.
• Agentic Capabilities: Enhanced through Multi-Teacher On-Policy Distillation (MOPD) and large-scale agentic RL during post-training, the model demonstrates superior tool-use capabilities and exceptional performance on benchmarks like SWE-Bench.
• Inference Efficiency: Pre-trained using FP8 mixed precision, making it highly optimized for practical deployments and modern accelerators.

For more details, please refer to the technical report, and the official repository.
This model was contributed by casinca.

Usage examples

Text generation

The example below demonstrates how to generate text with [Pipeline] or the [AutoModelForCausalLM] class.

import torch
from transformers import pipeline

pipe = pipeline(
    task="text-generation",
    model="XiaomiMiMo/MiMo-V2-Flash",
)
pipe("Explain why sparse MoE models can be efficient at inference.")

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("XiaomiMiMo/MiMo-V2-Flash")
model = AutoModelForCausalLM.from_pretrained(
    "XiaomiMiMo/MiMo-V2-Flash",
    device_map="auto",
)
input_ids = tokenizer("Explain why sparse MoE models can be efficient at inference.", return_tensors="pt").to(model.device)

output = model.generate(**input_ids, max_new_tokens=128)
print(tokenizer.decode(output[0], skip_special_tokens=True))

Chat template generation

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "XiaomiMiMo/MiMo-V2-Flash"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
)

messages = [
    {"role": "system", "content": "You are MiMo, a helpful assistant."},
    {"role": "user", "content": "Write a short summary of MiMo-V2-Flash."},
]

input_ids = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    return_tensors="pt",
).to(model.device)

generated_ids = model.generate(input_ids, max_new_tokens=128)
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)

MiMoV2FlashConfig

autodoc MiMoV2FlashConfig

MiMoV2FlashModel

autodoc MiMoV2FlashModel - forward

MiMoV2FlashForCausalLM

autodoc MiMoV2FlashForCausalLM - forward