Terminal-Bench 4.0.0 · GPT-6 Sol / Codex · reasoning effort max
PublicRerun job
330 / 330 finished
Started Sep 29, 2026, 5:33 PM · Ran for 11h 55m
328 completed2 timeouts0 errors0 in progress0 pending
Average reward
0.50
Across scored trials
Total tokens
5,448,732,388
16,511,310 tokens/trial · 99% cache hit
Inference cost
$1,597.29
$4.84 per finished trial
Sandbox cost
—
Results
Metric: reward
Group by0 groups
Loading results…