Xiaomi's open-source MiMo V2.6 Pro hits 46 on AA index at $0.13/task, near GPT-5.6 Sol
LearnWithBishal · x · 2026-09-23
Xiaomi's new open-source MiMo V2.6 Pro scored 46 on the Artificial Analysis Intelligence Index, making it the top-ranked open-source model — just 1 point behind GPT-5.6 Sol and ahead of Qwen3.8-Max, Kimi K3, and GLM-5.3.
On cost it runs about $0.13 per index task; Grok 4.7, which launched the same day with the same 46 score, costs $2.73/task at high reasoning and $3.74/task at xhigh — roughly 20x more.
The fully multimodal model targets complex research, long-horizon reasoning, scientific workflows, and cybersecurity tasks. Xiaomi attributes the gains to a post-training self-improvement system rather than changes to the pre-training foundation. Benchmarks look strong, but real-world developer experience remains to be seen.
More from Models
- Leak Claims OpenAI Will Launch Entire GPT-6.2 Family Tomorrow to Counter Opus 5.5 and Fable 5.2 — heypearlai · 2026-11-12
- ARC-AGI-3 leaderboard: models score 90%+ but at $10K+ per task — haider1 · 2026-11-12
- LMArena Launches Agent Arena for Head-to-Head Agent Evaluation — arena · 2026-11-12
- LLMs lack hyperlocal municipal data, explains disputed Texas power outage debate — csuwildcat · 2026-11-12
- Claude API's per-message effort lets you switch reasoning levels without breaking prompt cache — lyraxana · 2026-11-12
- rasbt coins "Benchmaxxing": making the numbers look better (best optics) — infoxiao · 2026-11-12