GPT-5 (minimal)

openai/gpt-5-20250807

Every ranked configuration of this model. Rank is dollars per unit of score on that job, and the unit is the job's own — the `per` column names it, because a rating point and a solved task are not the same thing to divide by. Displayed values are rounded; exact values in site.json and the dumps.
job	benchmark	#	score	task $	ratio $	per	released
Code	ALE-Bench	21	807.65	$0.043744	$0.00005416	rating point	2025-08-07
Code	ALE-Bench	40	1162.45	$0.244948	$0.0002107	rating point	2025-08-07
Reason	ProofBench	25	0.18	$0.852361	$4.735	correct answer	2025-08-07

These are the numbers on #frontier. How a fixed capability's cost falls over time is on /methodology.