
What happened
Xiaomi released and open-sourced the MiMo-V2.6 series — MiMo-V2.6-Pro and a smaller Flash variant — plus a Pro-UltraSpeed variant up to 20 times faster. Pro scored 46.32 on the Artificial Analysis Intelligence Index.
Why it matters
Xiaomi says Pro is the highest-ranked open-source/open-weight model in that comparison at launch, with 1 million-token context windows, and it claims Pro costs roughly one-20th to one-60th as much as overseas models at comparable intelligence.
What to watch
Pro still trails Claude Fable 5.1 and GPT-6 Astra on aggregate measures, and Xiaomi says the relevant business comparison is price per task, not price per token. Watch whether 20× faster generation justifies Pro-UltraSpeed's $4.35 and $8.70 pricing.
WHO IT HITSIt lands on teams evaluating cheaper open-weight alternatives for computer-use agents, coding and design agents, and document-heavy workflows, since the series is open-sourced and priced at MiMo-V2.5 API rates.
Summaries like this, every morning.
Xiaomi's move continues a pattern of open-weight releases aimed at agentic use cases. The company demonstrated the model on game worlds, Blender-based 3D modeling, embodied simulation and video-music workflows, and highlighted computer-use agents that read a task, a screenshot or video, reason, and decide the next step within one model loop.
On agentic tasks, V2.6-Pro's testing results were close to or occasionally above leading closed systems: 53.1 on AutomationBench versus Claude Opus 5's 50.3; 31.6 on Agents' Last Exam, tying Opus 5; and 89.9 on Terminal Bench 2.1, slightly ahead of Opus 5's 89.1. Yet the model still trails Claude Fable 5.1 and GPT-6 Astra on aggregate measures.
The pricing argument is Xiaomi's own: it claims Pro costs roughly one-20th to one-60th as much as overseas models at comparable intelligence, factoring in cached-token costs, versus $10 to $50 input and output per million tokens for GPT-6 Astra and Claude Fable 5.1. Whether that translates into lower cost per task is likely to hinge on reliability and how many loops a task takes to complete successfully.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
OpenAI said Apollo Research, METR, Transluce and Goodfire will work on-site at its offices, with access to unr…

Walt Disney Imagineering R&D revealed three Audio-Animatronic prototypes—Vilo Vele, Kado Bre'a, and Jibu Bre'a…

Citibank changed its Online User Agreement this week so customers who grant third-party AI tools access to the…

Google rolled out temporary threads in AI Mode, letting users click an icon in the upper right corner to chat…

OpenAI CEO Sam Altman said at the New York Times DealBook Summit that he had spoken privately with former Nvid…

Reson8 released Rumble-2-Flash, a 500 million-parameter transcription model supporting over 40 languages, as a…