HarnessOpt-Bench - Terminal-Bench: leaderboard

Metric: Normalized gain ((held-out score minus seed score) / (1 minus seed score); negative when the nominated harness is worse than the seed; mean over the contestant's optimization rounds, each held-out case scored three times; Terminal-Bench 2.0 terminal tasks with a grok-build target agent; seed held-out score 0.241, 36 test cases; the optimizer edits the seed harness under fixed development and validation evaluation budgets and never sees the test partition). Source: arxiv.org. Saturation forecast: Around May 2028. 10 models tracked.

Top models

#ModelScore
1Claude Opus 50.29
2Claude Sonnet 50.15
3GPT-5.6 Sol0.13
4Kimi K30.12
5GPT-5.6 Terra0.04

Interactive version: theaggregate.ai/benchmark?slug=harnessopt-bench-terminal-bench · How It Works · Data refreshed daily, snapshot 2026-09-29.