goose evaluates five open-source coding harnesses: similar pass rates, with per-task costs differing by up to 75%
Same Brain, Different Bill
AI BriefingThe official goose team used the same model, gpt-6.1-sol (high reasoning effort), and 62 tasks from Terminal-Bench 4.0 to conduct 310 scoring trials each for five open-source harnesses: goose, Codex CLI, OpenCode, Hermes Agent, and pi.
Source: goose official blog · goose-docs.ai