MiMo-7B
MIT-licensed 7B open reasoning model that punches above its weight in math and code.
About
Xiaomi's first open model, released April 2025, is a 7B reasoning model family (base, SFT, RL-Zero, RL, and the later RL-0530 checkpoint) under the MIT license — genuinely free for commercial use, no strings. Xiaomi reports the RL variant hitting 95.8% on MATH-500 and 57.8% on LiveCodeBench v5, which is remarkable for the size, and it runs comfortably on a single consumer GPU.
It is a narrow specialist: math and code reasoning are the point, and general knowledge scores sit in the mid-50s. The team has since moved on to the much larger MiMo-V2 line, so 7B is a finished artifact, not a living family.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
- MATH-500: 95.8%
- AIME 2024: 68%+
- LiveCodeBench v5: 57.8% (all Xiaomi-reported)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: MIT license confirmed across multiple sources; genuinely commercial-friendly; Released 2025-04-30; RL-0530 checkpoint refreshed May 2025; Team has moved to MiMo-V2 line — 7B is a finished artifact; Benchmarks are Xiaomi-reported. License: MIT confirmed from multiple sources including Wikipedia's Xiaomi MiMo page..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 7.0/10
- Maintenance · recency, releases, issue load
- 6.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · MIT license confirmed across multiple sources; genuinely commercial-friendly
- · Released 2025-04-30; RL-0530 checkpoint refreshed May 2025
- · Team has moved to MiMo-V2 line — 7B is a finished artifact
- · Benchmarks are Xiaomi-reported