Thursday22 Oct 2026
Artificial Analysis launches Coding Agent Index v2.0 across 290 tasks
The updated index weights Cognition’s FrontierCode v1.1, Laude Institute’s Terminal-Bench 4.0 and Scale AI’s revised SWE-Atlas-QnA-AA equally, with Claude Opus 5.5 in Claude Code leading at 61.7.
Coding agentStory 11 of the day
