GPT-5.6 Sol (max)

openai/gpt-5.6-sol

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	62	2176.88	$1.53408	$0.0007047	rating point	2026-07-09
Reason	ProofBench	13	0.77	$1.7366	$2.255	correct answer	2026-07-09
Agent	DeepSWE	6	0.454	$1.07427	$2.369	solved task	2026-07-09
Agent	DeepSWE	9	0.611	$1.86203	$3.049	solved task	2026-07-09
Agent	DeepSWE	16	0.694	$3.46983	$5	solved task	2026-07-09
Agent	DeepSWE	20	0.707	$4.70366	$6.65	solved task	2026-07-09
Agent	DeepSWE	32	0.727	$8.38644	$11.54	solved task	2026-07-09

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