MiMo-VL
XiaomiMiMo-VL is Xiaomi's 7B vision-language model line, joining a native-resolution vision encoder to the MiMo-7B language model. It answers questions about images, documents and video and reasons step by step, with a switch that turns the long reasoning off. The 2508 release, the current one, ships as SFT and RL checkpoints and raised the line's image and video benchmark scores.
The repository's LICENSE is Apache-2.0 while the model cards say MIT; both are permissive, and the cards govern the weights.
Openness
3 high confidence- weights
- open (MiMo-VL-7B-RL-2508 and SFT-2508 on the Hub, ungated)
- data
- described (pretraining and RL data described in the technical report, not released)
- code
- partial (inference demo and notebooks only)
- license
- MIT (weights)+Apache-2.0 (repository code)
The 2508 checkpoints are MIT and the repository code is Apache-2.0. Xiaomi describes its training stages and data in the report but publishes neither the data nor the training code, only inference examples.
- https://huggingface.co/api/models?author=XiaomiMiMo&limit=100&expand[]=downloads&expand[]=cardData&expand[]=gated&expand[]=createdAt recorded 2026-11-12
MiMo-VL-7B-RL-2508 and MiMo-VL-7B-SFT-2508 (2025-08-07), the newest MiMo-VL checkpoints, gated false and tagged mit.
- https://huggingface.co/XiaomiMiMo/MiMo-VL-7B-RL-2508/raw/main/README.md recorded 2026-11-12
MiMo-VL-7B-RL-2508 card: license mit; "Both versions of the MiMo-VL-7B-2508 model are now open-sourced on Hugging Face"; a use-restrictions notice that states it "does not amend the applicable model licence".
- https://raw.githubusercontent.com/XiaomiMiMo/MiMo-VL/main/LICENSE recorded 2026-11-12
LICENSE: Apache License, Version 2.0.
- https://raw.githubusercontent.com/XiaomiMiMo/MiMo-VL/main/README.md recorded 2026-11-12
README: a four-stage pretraining phase and Mixed On-policy Reinforcement Learning described; "We open-source MiMo-VL-7B series, including checkpoints of the SFT and RL model"; no dataset release.
- https://ungh.cc/repos/XiaomiMiMo/MiMo-VL/files/main recorded 2026-11-12
Repository tree: demo/app.py, demo/infer.py and example notebooks for grounding and OCR; no training code.
Adoption
2 medium confidenceHugging Face downloads over the trailing 30 days for Xiaomi's six MiMo-VL checkpoints, including its own GGUF builds, most of them for the 2508 RL model.
- https://huggingface.co/api/models?author=XiaomiMiMo&limit=100&expand[]=downloads&expand[]=cardData&expand[]=gated&expand[]=createdAt recorded 2026-11-12
MiMo-VL-7B-RL-2508 4,587, SFT-2508 1,978, RL 1,581, SFT 1,017, RL-GGUF 621 and SFT-GGUF 349 downloads in the trailing 30 days (10,133 together).
Capability
3 medium confidenceMiMo-VL reads images, documents and video, level with MiniCPM-V. Xiaomi reports it locating elements on a screen but gives no results for completing tasks in a live interface.
- https://huggingface.co/XiaomiMiMo/MiMo-VL-7B-RL-2508/raw/main/README.md recorded 2026-11-12
Card: "consistent improvements across both image and video benchmarks, achieving notable milestones of 70.6 on MMMU and 70.8 on VideoMME"; thinking control through /no_think.
- https://raw.githubusercontent.com/XiaomiMiMo/MiMo-VL/main/README.md recorded 2026-11-12
README: "MiMo-VL-7B-RL possess exceptional GUI understanding and grounding capabilities"; video question-answering example; OCR and grounding notebooks.
Verified 2026-11-12