Coding Agent Index v2.0 launches: Claude Opus 5.5 leads at 61.7, Max mode results disputed
ryanallred · x · 2026-10-23
Artificial Analysis shipped Coding Agent Index v2.0, evaluating models together with their agent harnesses across 290 tasks equally weighting software engineering, terminal use, and repo Q&A. Cognition's FrontierCode v1.1 (100 private real-repo tasks, graded on maintainer merge-worthiness) replaces DeepSWE, SWE-Atlas-QnA grading was revised, and Claude Opus 5.5 in Claude Code leads at 61.7. Critics argue Max mode results should be removed since they inflate scores and tokens per task, misleading users.
More from coding & agent
- Devs on Reddit swap tactics for stopping prompt regressions in AI agents — SleepFlimsy6963 · 2026-10-23
- Claude Code + ComfyUI MCP workflow for client work: paid models or open weights? — newaccount47 · 2026-10-23
- Social scientist: CS treats agentic development as a magic black box — RexDouglass · 2026-10-23
- Someone vibe-coded a Tesla Roadster game — before GTA VI — Jasonio · 2026-10-23
- Codex 0.164.0 'account/read failed' TUI error — fix is restarting the app-server daemon — bytebot · 2026-10-23
- Same native UI, two stacks: Swift build is ~80MB Mac-only, Zig build is ~10MB cross-platform — jedisct1 · 2026-10-23