Atlas

Benchmarks

← All benchmarks

RuneBench Gold - Fletch + Alch 30m

Games · 2026-02-26

RuneBench Fletch + Alch 30m gold-accumulation task. Agents start from a fletching and alchemy-oriented condition and are scored by peak coins seen during the run.

Top models (higher is better)

ModelScore
Opus 4.624467
Opus 4.722200
Sonnet 4.611049
Opus 4.56564
Sonnet 4.524.0
Haiku 4.53.0
Loading Atlas data…