Skip to content

Sign in with GitHub

Terminal-Bench 2.0 on Claude Sonnet 4.5

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

Build a run for this board

Pass rate by harness, higher is better

HarnessPass rate

final1.5-official42.7%95% interval 32.6–52.8

All harnesses

Reading this table

Every harness on this board, in its order
Harness	pass^5	Runs	Last published
final1.5-official	42.7%	13 infra	25 Oct 2026run #59

Every result is self-reported
Data as of 13 Nov 2026
Methodology
API
Report a submission