Terminal-Bench 2.0 on GPT-5.5

89 tasks, 5 attempts, open network. Updated 25 Oct 2026.

Pass rate by harness, higher is better

Reference

Δ vs Capy, pts

Capy83.1%95% interval 76.6–89.2Capy is the reference.

clnkr0.3.1166.1%95% interval 57.5–74.2clnkr 0.3.11 minus Capy: −17.1 points, 95% interval −24.5 to −10.1, over 89 tasks both ran; a clear difference.

Every harness on this board, in its order
Harness	Rank	pass^5	Last published
Capy	1	69.7%	25 Oct 2026run #47
clnkr0.3.11	2	50.6%	25 Oct 2026run #89