OpenAI usage limits have been severely cut, and intelligence appears to be markedly declining, so I'm going to start trying these Chinese models seriously now. I don't mind if it takes longer. I just need the intelligence to predictably work the same way from day to day.
It feels suspicious that MiMo-V2.6
Pro gets 46 in de index while DeepSeek-V4.1 (https://artificialanalysis.ai/models/deepseek-v4-1-flash) gets 39. According to the appendix at the bottom of https://mimo.xiaomi.com/mimo-v2-6 the deepseek model sometimes surpasses mimo and it's not so far behind in capabilities. A week ago opus 5 appeared 1 points ahead of fable 5 despite fable being a much smarter model (this has been corrected already)
It is an impressive model. Agreed on most that is written on this page, with the exception of it being fast. I ran it on my own LLM benchmark suite[1] and it is faster than DeepSeek but still much slower than leading models. But it's pricing is where it really shines.
KillSwitch-Bench 1.0
Claude Opus 5 66.9
GPT-6 Astra 57.9
Claude Fable 5.1 46.7
MiMo-V2.6-Pro 38.8
Muse Spark 1.3 36.5
Per Xiaomi, MiMo v2.6 training run cost $3.47m. A far cry from the estimated costs ($100m+) for the Big 5 (MSL, xAI, GDM, OAI, Ant). I wouldn't be surprised if salaries and R&D costs have similar drastic disparities.
For a model that matches Muse Spark 1.3 in benchmarks, MiMo v2.6 Pro is incredibly cheap, given its cache rates will remain $0.0036 per million.
MiMo-v2.6-Pro: Intelligence, Performance and Price Analysis
(artificialanalysis.ai)68 points by theanonymousone 6 hours ago | 19 comments
Comments
KillSwitch-Bench 1.0
1 - https://bench.killswitch-lang.org/For a model that matches Muse Spark 1.3 in benchmarks, MiMo v2.6 Pro is incredibly cheap, given its cache rates will remain $0.0036 per million.