Xiaomi's affordable flagship AI leads the open models, and Anthropic says Claude helped get it there
Xiaomi's MiMo-V2.6-Pro tops open-model rankings at a fraction of rivals' cost, while Anthropic accuses Xiaomi of large-scale Claude distillation.
Xiaomi released MiMo-V2.6-Pro, a 1.02-trillion-parameter MoE with 42B active, scoring 46 on Artificial Analysis's Intelligence Index to lead open models ahead of Kimi K3 and Qwen, priced at $0.435/M input and $0.87/M output tokens. Performance gains came from a scaled reinforcement-learning run costing $2.62M in under six days, lifting DeepSWE from 58.4 to 72.6. Anthropic's threat report (case GTG-16008) tracks over 400,000 exchanges routing MiMo conversations through OpenClaw and OpenCode to Claude for illegal distillation. Xiaomi also open-sourced its RL toolkit with roughly 7,000 auto-graded training tasks, including cybersecurity tasks drawn from OSS-Fuzz.
- MiMo-V2.6-Pro: 1.02T-parameter MoE, 42B active, tops open-model Intelligence Index at 46
- Costs $0.435/M input and $0.87/M output tokens; RL training run cost $2.62M in six days
- Anthropic accuses Xiaomi of distilling Claude via 400,000+ exchanges (case GTG-16008)
- Open-sourced RL framework with ~7,000 auto-graded tasks including OSS-Fuzz cybersecurity tasks
- Organizations
- XiaomiAnthropicMiMo-V2.6-ProArtificial Analysis
- AI models
- Claude
From the original article
3 min read at the-decoder.com
With MiMo-V2. 6-Pro, Xiaomi moves to the top of the openly available AI models and drastically undercuts the competition on price. What makes that possible is massive reinforcement learning that cost $2.
By Maximilian Schreiner · the-decoder.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.