Kimi K3 Ties for 5th in Coding, Beats GPT-5.6 in Value
scaling01 · x · 2026-07-18
In the Artificial Analysis Coding Agent Index, Kimi K3 scored 57, tying for 5th place with models like GPT-5.6 Terra and leading Opus 4.8. In specific evaluations, it delivered outstanding results in Terminal-Bench, DeepSWE, and others. Furthermore, the average cost per task for K3 is only $3.18—55% cheaper than GPT-5.6 Sol max—demonstrating exceptional cost-effectiveness for frontier coding.
More from Models
- Cohere's Invent-A-Dataset paper: zero-seed synthetic data beats frontier APIs on quality and diversity — idanbeck · 2026-10-27
- ARC Prize redesigns leaderboard: highest reasoning effort by default, CSV export — mhmazur · 2026-10-27
- Cohere: open-weight Qwen3.8-Flash-Next tried uploading probe file to Beijing cloud bucket in sandbox — wstone_bd · 2026-10-27
- Grok Bot auto-connects to X Money accounts without opt-in, user pulls funds — noahsolomon · 2026-10-27
- Model refuses wrong psychiatric advice in deliberately misleading test — FellMentKE · 2026-10-27
- Clinical System One v0.2: top clinical AI drops from 97% to 65% on harder cases — MaziyarPanahi · 2026-10-27