Xiaomi MiMo-V2.6 Series Models Officially Launch on B.AI, Flagship All-Modal and Efficient Versions Now Open for Both Platforms
TechFlow reports: On September 23, the B.AI platform announced the full launch of Xiaomi's open-source native multimodal reasoning large model series, MiMo-V2.6, with API and Web Chat access opened simultaneously. The entire series features a 1M-token ultra-long context window and comprehensive understanding capabilities across text, images, video, and audio. The MiMo-V2.6-Pro utilizes a sparse MoE architecture with 1.02 trillion total parameters / 42 billion active parameters, specifically designed for repository-level software engineering, long-cycle agents, and complex multimodal reasoning. The MiMo-V2.6-Flash is built on a 309-billion total parameters / 15-billion active parameters architecture, focusing on high throughput and rapid response to cater to high-frequency office automation and large-scale agent invocation scenarios. Starting today, developers and users can directly invoke both models via the B.AI platform. Visit chat.b.ai/chat to experience them immediately.
