Xiaomi's MiMo v2.6 sets a new bar for open-model transparency
Xiaomi MiMo v2.6
Xiaomi's MiMo v2.6, a 9B agentic model built by fine-tuning Qwen3.5-9B on MiMo-generated data, has sparked praise for its unusually transparent release. The company shared a real-time training dashboard, a detailed tech report, and benchmark scores including failures. Commenters highlight its cost-effectiveness and strong performance on niche tasks, with some calling it a Pareto frontier model for game-playing under $0.15 per million input tokens.
If you’re releasing an open model going forward, please consider offering the community more of this transparency!
- rao-v
I know we have strong views on what a truly open model is (open weights, open training data, open training code etc.) but I really like how transparent they’ve been about the training of this model.
The realtime dashboard they shared during training (https://mimo.xiaomi.com/rl/) was an incredible learning and teaching tool for me, and they’ve been unusually comprehensive in sharing details about their methodology (check out that tech report - it's got lots of clever behind the scene tricks like Google or Deepseek writeups) and benchmark scores (even the stuff they didn’t do well on).
If you’re releasing an open model going forward, please consider offering the community more of this transparency!
- lwansbrough
Anyone else more excited about Chinese models than American models these days? Big thing for me is affordability.
- stymaar
Flash[1]: 309B total / 15B activated parameters
Pro [2]:, 1.02T total / 42B activated parameters
- simonw
Pelicans for Flash: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
Pelicans for Pro: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
- margorczynski
China will most probably win the AI race in the long run because of one major bottleneck the US has - energy. The electric energy and grid buildout in China has been massive since a long time and there is simply no way for the US to quickly catch up.
No matter how much cash you throw you can't just materialize a 100 nuclear reactors to power the data centers.
- user43928
I don't trust any of the benchmarks where Opus 5 surpasses Astra or Fable 5.1.
Maybe Terminal Bench 4.0 and ExploitGym are reasonable.
Terminal Bench 4.0
GPT 6 Astra 59.6
Claude Fable 5.1 55.1
Claude Opus 5 49.0
MiMo-V2.6-Pro 34.9
MiMo-V2.6-Flash 28.8
DeepSeek V4.1 Flash 26.8
MiMo-V2.5-Pro 1.5
ExploitGym
GPT 6 Astra 42.4
Claude Fable 5.1 30.4
Claude Opus 5 22.1
MiMo-V2.6-Pro 17.8
MiMo-V2.6-Flash 6.0
MiMo-V2.5-Pro 0.1
DeepSWE v1.1
DeepSeek V4.1 Flash 74.2
Claude Opus 5 74.0
GPT 6 Astra 74.0
MiMo-V2.6-Pro 71.9
Claude Fable 5 70.0
MiMo-V2.6-Flash 67.9
MiMo-V2.5-Pro 19.0
- nemothekid
Looking at the frontend design examples; why do these models seem to love the "01 - UPPERCASE TEXT" motif. It's everywhere now (see https://try.cloudflare.com/, which has '01 · QUICK TUNNELS', but no "02" anywhere).
- vatsachak
Wow, the chinese labs are getting good at advertising model releases. The moat is thin.
Some features of the release I like:
- Demonstration of diverse tasks, such as using a DAW
- Graphs from various benchmarks and price ranges
- Real world use of the model in scientific environments
- toephu2
I said this years ago, LLMs are a commodity (or were becoming one at the time). They are dime a dozen. Even the frontier ones. OpenAI and Anthropic have no moat.
No moat and competition is good for consumers though.
- paradox460
Mimo has long been one of my preferred models. It works well at many code tasks, generally has a pleasant voice, and isn't prone to over-analzing and researching
Also when I was using it, I managed to do quite a bit on a few bucks worth of OpenRouter credits. Not sure how well it keeps up in the modern world against things like Luna, but I hope it remains competitive