5 tracked versions of the mimo-v2 family. Timeline ordered newest → oldest. Flagship version marked.
MiMo-V2.5 is a native omnimodal model by Xiaomi. It delivers Pro-level agentic performance at roughly half the inference cost, while surpassing MiMo-V2-Omni in multimodal perception across image and video understanding...
MiMo-V2.5-Pro is Xiaomi’s flagship model, delivering strong performance in general agentic capabilities, complex software engineering, and long-horizon tasks, with top rankings on benchmarks such as ClawEval, GDPVal, and SWE-bench Pro....
MiMo-V2.6-Distill-Qwen-9B is Xiaomi's 9B supervised fine-tuning checkpoint built from Qwen3.5-9B using MiMo-generated agent data. It targets coding, tool use and visual coding; it is separate from the hosted MiMo Pro and Flash models.
MiMo-V2.6-Flash is Xiaomi's lower-priced multimodal API option in the September 22 release. Its downloadable MOPD checkpoint has an MIT license. The published token rate is useful for screening high-volume workflows; it does not establish cost per successful task or a latency guarantee.
Xiaomi's September 22 multimodal reasoning model, available through the MiMo API and downloadable MIT-licensed MOPD weights. The standard API and the separately priced UltraSpeed route require different budgets. Evaluate tool-call behavior on your own tasks before migrating.
See every provider hosting Xiaomi models at /provider/xiaomi.
Fast Pivot — Access hundreds of AI models from a single API — competitive pricing, reliable uptime, enterprise-grade infrastructure.