---
title: "Xiaomi's MiMo-V2.5 Matches Claude and Gemini With One Open MIT Model"
subtitle: "Xiaomi's open MIT-licensed omnimodal MoE handles text, image, video, and audio in one 1M-token model, activating just 15B of 310B params."
url: "https://alphasignal.ai/news/xiaomi-s-mimo-v2-5-matches-claude-and-gemini-with-one-open-mit-model"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-04-27T13:37:38.000Z"
updated: "2026-08-22T02:13:22.419Z"
type: "Model"
topics:
  - "Llms"
  - "Agents"
subtopics:
  - "Long Context"
  - "Vision Language"
  - "Computer Use"
upvotes: 3693
source_url: "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"
comments: 0
truncated: true
---

# Xiaomi's MiMo-V2.5 Matches Claude and Gemini With One Open MIT Model

> Xiaomi's open MIT-licensed omnimodal MoE handles text, image, video, and audio in one 1M-token model, activating just 15B of 310B params.

Xiaomi has released [MiMo-V2.5](https://huggingface.co/XiaomiMiMo/MiMo-V2.5), an open-weights omnimodal model that reads text, images, video, and audio through a single unified architecture. It ships with a 1M token context window, a Mixture of Experts backbone, and benchmark numbers that put it in the same conversation as Claude Opus and Gemini 3 Pro, all under an MIT license.

Rather than stitching together a vision model, an audio model, and an LLM, everything runs through one network with dedicated encoders feeding a shared MoE backbone. Xiaomi also positions it as agentic, with tool-calling and long-horizon task performance as first-class targets.

## Under the hood

The base MiMo-V2.5 is a sparse MoE with 310B total parameters and 15B activated per token, drawing from a pool of 256 routed experts with 8 selected per token. A heftier [MiMo-V2.5-Pro](https://mimo.xiaomi.com/mimo-v2-5/) variant weighs in at 1.02T total and 42B activated parameters for teams that want the biggest brain available. Both share the design inherited from the MiMo-V2-Flash backbone.

The attention mechanism is where things get interesting. It interleaves Sliding Window Attention and Global Attention in a 5:1 ratio with a 128 token window, which the team reports cuts KV-cache storage by nearly 6x, while a learnable attention sink bias preserves long-context recall. That trick is what makes the 1M context window financially viable to serve.

The perception stack is purpose-built rather than bolted on:

* A 729M parameter Vision Transformer with 28 layers (24 sliding window, 4 full attention) using a hybrid row and column window pattern

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [XiaomiMiMo/MiMo-V2.5 Model Card](https://huggingface.co/XiaomiMiMo/MiMo-V2.5)
- [MiMo-V2.5 Official Announcement](https://mimo.xiaomi.com/mimo-v2-5/)
- [Xiaomi MiMo-V2.5 on OpenRouter](https://openrouter.ai/xiaomi/mimo-v2.5)
- [Xiaomi rolls out MiMo V2.5 with multimodal AI and improved efficiency](https://crypto.news/xiaomi-rolls-out-mimo-v2-5-with-multimodal-ai-and-improved-efficiency/)
- [Xiaomi releases open-weight MiMo-V2.5 AI model](https://www.gsmarena.com/xiaomi_releases_openweight_mimov25_ai_model_claims_frontierlevel_agentic_capability-news-72585.php)