Xiaomi MiMo-V2.6-Pro tops open-weights AI index by slim margin, $6 plan has credit caveats

Xiaomi's MiMo-V2.6-Pro scored 46 on the Artificial Analysis Intelligence Index v4.3, making it the highest-ranked open-weights model currently available, edging out GLM-5.3 and Kimi K3 by just one and two points respectively. The model is open-source under an MIT licence, though viral posts overstated the performance gap and incorrectly identified the Kimi competitor as K2 rather than K3. In a coding benchmark by independent outlet kingy.ai, MiMo-V2.6-Pro matched Claude Opus 5 on all 23 test cases but at roughly one-thirtieth of the cost per run, helping explain widespread excitement. Xiaomi's $6-per-month Lite plan offers 4.1 billion credits with no hourly usage caps, but credits are not equivalent to tokens — context-heavy tasks using the Pro model can consume two to four credits per token, causing faster depletion than many users expect. The model is a Mixture-of-Experts architecture with 1.02 trillion total parameters, 42 billion active, and supports up to one million tokens of context.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.


Discussion (0)
Log in to join the discussion and vote.
Log in