Latest

An MIT licence removes the commercial-use restrictions that still accompany many open-weight releases, so teams can fine-tune or ship it in products without negotiating terms. The hybrid SWA–DSA design without any full-attention layers is a notable architectural signal for anyone tracking how long-context efficiency is evolving beyond dense attention. Engineering teams evaluating open-weight alternatives for code generation and research pipelines should benchmark it against comparably sized MoE releases.

官方一手pandaily@[email protected] (Pandaily)Source

The reveal signals Xiaomi is positioning MiMo-V3 specifically for agent workloads, where attention efficiency over growing histories matters more than raw benchmark scores. Engineers building multi-turn agents can compare HySparse2's KV sharing against Alibaba Qwen and DeepSeek approaches for long-context cost control. The architecture details also hint at what RL training changes shipped in MiMo-V2.6.

The architecture targets agentic workloads, where each turn appends long observations and prefill cost, cache size and retrieval accuracy sit on the critical path simultaneously. Switching from block-level to token-level selection and folding SWA into a forced recent window gives local and global tokens one shared KV cache. Developers building long-context agents should watch MiMo-V3's release, especially against open-source alternatives like DeepSeek and Qwen in the same weight class.

For users interested in large language models, Xiaomi's MiMo-V2.6 is a significant update that demonstrates advancements in reinforcement learning and multi-task capabilities.

For users looking for affordable AI models, mimo-v2.5's strong performance on OpenRouter and mimo-v2.6's improved results show Xiaomi's models remain competitive in price and capability compared to alternatives like GLM-5.3-Flash.

Blognone@lew2 sourcesSource

小米新出的语音模型 MiMo-V2.5-ASR 能听方言、混中英文、扛噪音，每小时才五毛钱，适合做语音转录的朋友试试。

1万亿参数跑出1000t/s

官方一手marktechpost@Asif RazzaqSource