Kimi K3.1 Pro posts 67% on Terminal-Bench 4.0, highest of any model
ArtificialAnlys · x · 2026-10-24
Artificial Analysis details Kimi K3.1 Pro's agentic results: 3rd on GDPval-AA with 1,846 Elo (+312 vs Kimi K3), behind only Claude Opus 5.5 (1,866) and Muse Wonder Full (1,857); 67% on Terminal-Bench 4.0 (up from 13%), the highest of any model and ahead of Claude Sonnet 5.5 (64%); AutomationBench-AA for agentic SaaS workflows rises from 58% to 73%.
Related event: Kimi K3.1 Pro Ranks 4th on AA Intelligence Index with Major Agent Gains(9 posts)→
More from Models
- Kimi K3.1 Pro claimed to near Claude Opus coding, free download on Nov 4 — prasenx · 2026-10-24
- Labs are far ahead of what they release — and the gap may be widening, speculates X user — MoonL88537 · 2026-10-24
- API Pricing Cut from $15 to $10 per Million Tokens — cedric_chee · 2026-10-24
- MiniMax's 430B model beats 3T-scale rivals on AA knowledge benchmark — teortaxesTex · 2026-10-24
- Hand-crafted regex pretokenizers quietly do heavy lifting in every frontier LLM — YouJiacheng · 2026-10-24
- Block Unveils Dessa, a Full-Duplex Speech-Only Model Trained From Scratch With No Text — stochasticchasm · 2026-10-24