Open-source (Apache 2.0) LLM series 'born for reasoning.' Pre-trained & RL-tuned models (like the 7B) match o1-mini on math/code. Base/SFT/RL models released.
Viktor.comAn AI coworker that actually does the work
MiMo Launches
MiMo-V2-Pro & OmniXiaomi's flagship agentic and omni-modal foundation models
Launched on March 19th, 2026
