Claude Sonnet 5 (high)

anthropic/claude-sonnet-5

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	61	1463.12	$0.96046	$0.0006564	rating point	2026-06-30
Reason	ProofBench	21	0.66	$2.24876	$3.407	correct answer	2026-06-30
Agent	DeepSWE	23	0.305	$2.18656	$7.166	solved task	2026-06-30
Agent	DeepSWE	29	0.398	$4.07904	$10.25	solved task	2026-06-30
Agent	DeepSWE	36	0.482	$7.42556	$15.4	solved task	2026-06-30
Agent	DeepSWE	42	0.497	$11.8906	$23.94	solved task	2026-06-30
Agent	DeepSWE	44	0.538	$26.3999	$49.03	solved task	2026-06-30

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