Kimi K3.1 Pro Scores 65.7% on Terminal-Bench 4.0 at Roughly a Quarter of Opus 5.5's Cost
Kimi_Moonshot · x · 2026-10-24
Kimi's thread adds benchmark detail: K3.1 Pro scores 65.7% on Terminal-Bench 4.0, which evaluates coding agents on realistic, resource-constrained terminal tasks — ahead of GPT-6 Astra (58.2%) and narrowly behind Claude Opus 5.5 (66.4%). The cost-vs-score chart shows K3.1 Pro hits this at roughly $4-5 per task, versus about $15 for Opus 5.5.
Related event: Kimi K3.1 Pro tops multiple benchmarks, closing in on Claude Opus 5.5(7 posts)→
More from Models
- Kimi K3.1 Pro full results released on Artificial Analysis Intelligence Index — ArtificialAnlys · 2026-10-24
- Full evaluation charts for Kimi K3.1 Pro across the Intelligence Index — ArtificialAnlys · 2026-10-24
- Same size, 11 points smarter: Kimi K3.1 Pro tops the parameter-Pareto frontier — ArtificialAnlys · 2026-10-24
- Kimi K3.1 Pro's hallucination rate drops to 44%, lifting AA-Omniscience to 30 — ArtificialAnlys · 2026-10-24
- Kimi K3.1 Pro debuts 4th on Intelligence Index with big agentic gains, prices cut 33% — ArtificialAnlys · 2026-10-24
- Kimi K3.1 Pro posts 67% on Terminal-Bench 4.0, highest of any model — ArtificialAnlys · 2026-10-24