Xiaomi MiMo-V2.6 lands, jumping 153 pts to crack Code Arena top 10 among open weights

arena · x · 2026-09-22

Xiaomi releases MiMo-V2.6 (Pro & Flash)

• Two omnimodal models trained via scaled RL; weights, tech report, RL environments and training code released open-source under MIT
• The team claims Pro matches Claude Opus 5 and GPT-5.6 Sol on most agent benchmarks and scores 46 on the Artificial Analysis Intelligence Index — the highest among open models
• On Code Arena: WebDev it debuts at #10 overall (1628 AutoEval pts), #3 among open weights, just 7 pts behind Qwen3.8 Flash Next and 46 behind Kimi K3 Max — a +153 pt jump from MiMo-V2.5-Pro (1475 → 1628), tying Claude Fable 5 (High)
• Caveat: early AutoEval scores from a reward model, not final live human votes

Related event: Xiaomi open-sources MiMo-V2.6, topping open-weight leaderboard at 46 with livestreamed RL training(47 posts)→

Original post →

More from Models

• YC-backed Monte launches post-training platform to build proprietary models from your agents' real work — ycombinator · 2026-11-12
• 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

Models channel →