Xiaomi Deploys MiMo-V2.6-Pro to WorkBuddy with 1M Context and Multimodal Multi-Agent Support
Original title:小米 MiMo-V2.6-Pro 正式上线 WorkBuddy，支持 1M 超长上下文多智能体分工协作等
IT Home, October 20 — Xiaomi Technology officially announced today that Xiaomi MiMo-V2.6-Pro has officially launched on WorkBuddy. According to reports, Xiaomi MiMo-V2.6-Pro scored 46 on the Artificial Analysis Intelligence Index, ranking first among open-source models globally. It features a 1M ultra-long context window, multi-agent collaboration, and unified multimodal understanding across images, video, audio, and text.
As previously reported by IT Home, in the early hours of September 22, Xiaomi officially released and open-sourced the brand-new Xiaomi MiMo-V2.6 series models, which include two native multimodal models: Pro and Flash. Xiaomi stated that this is a crucial step in its exploration of the RSI (Recursive Self-Improvement) path, scaling up reinforcement learning compute to allow the models to expand the boundaries of intelligence through continuous exploration and feedback.
The MiMo-V2.6 series retains the API pricing of the V2.5 series: Flash is priced at 1 yuan per million input tokens and 2 yuan per million output tokens; Pro is priced at 3 yuan and 6 yuan, respectively, and offers a 99% cache discount. At an equivalent level of intelligence, the price is merely 1/20 to 1/60 of that of overseas models.
Why it's worth reading
It demonstrates how aggressive pricing and 1M-token context windows are making open multimodal models viable for high-frequency multi-agent enterprise tasks.