[AINews] Xiaomi MiMo-V2.6-Pro 1T-A42B: the new top Open Weights model, trained for $3M
Xiaomi released open-weights MiMo-V2.6-Pro, a 1.02T omnimodal model topping Artificial Analysis's open-model index.
Xiaomi released open-weight natively omnimodal models MiMo-V2.6-Pro and MiMo-V2.6-Flash, plus Pro-UltraSpeed mode claiming up to 20x faster generation at the same quality. Artificial Analysis ranks MiMo-V2.6-Pro first among open-weights models on its Intelligence Index with a score of 46; it has 1.02 trillion total parameters and 42 billion active, with cited pricing of $0.435 per million input tokens and $0.87 per million output tokens under an MIT license. Xiaomi described reinforcement learning across coding, agent, visual, and cyber tasks, with one cited run of about 130 hours, 75 billion tokens, and $2.6 million. Training recipes and environment code are slated for release, but the complete 7,000-plus task datasets are not yet public.