Xiaomi introduces Mimo-V2.6 series open-source AI model family
Global technology company Xiaomi Corp. today announced the release and open-sourcing of its MiMo-V2.6 series generative artificial intelligence models today, which includes two natively omnimodal models bringing a balance of intelligence, efficiency and cost. The MiMo-V2.6 series includes two natively omnimodal models with a new flagship model, MiMo-V2.6-Pro, and a smaller, efficient Flash…
Xiaomi Corp. has announced the open-sourcing of its Mimo-V2.6 series of generative AI models, which include two omnimodal models. These models feature a balance of intelligence, efficiency, and cost. The series includes a flagship model, MiMo-V2.6-Pro, a smaller and more efficient Flash variant, and a slimmed-down Pro-UltraSpeed variant with up to 20 times faster output compared to Pro.
The Pro and Flash models are described as omnimodal, accepting text, image, video, and audio inputs. The larger models have a 681 million-parameter vision encoder, a 308 million-parameter AudioTokenizer, and a 127 million-parameter audio patch encoder. This design allows them to handle agentic AI use cases and task instructions, particularly for computer use.
The Pro model scored 46.32 on the Artificial Analysis Intelligence Index, behind the most advanced proprietary closed-source models but ahead of other open-source models. Xiaomi claims the models cost significantly less than competitors, with the Pro model costing roughly one-20th to one-60th as much as GPT-6 Astra and Claude Fable 5.1.
Written by urgent.news from SiliconANGLE's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
