Skip to content

claude-sonnet-4-20250514

Anthropic

legacy model Its test configurations are retired, so it is no longer run. The figures are from its best legacy runs, scored as they ran at the time; it is not ranked.

average: 61.6
benchmarksof 12: 10
per 1,000 documentsmean over its benchmarks: $17
seconds per inputmean response time: 13.6

Compare with other models Its row in the Matrix All its test runs

By task

Data correction

61.0

Best of 114 models with all its benchmarks: claude-opus-5-5, 82.7.
Transcription

68.7

Best of 114 models with all its benchmarks: gemini-4-argon, 96.2.

By benchmark

Benchmark	Score	Rank	Per 1,000 docs	s / input	Best run
Bibliographic Dataolder run	95.9	–	$34$0.17 run	25.56	T0107 · 30 Sep 2025prompt.txt
Blacklist Cards	85.0±3.3	–	$11$0.35 run	5.37	T0323 · 24 Jan 2026prompt.txt
Book Advert XML Repair	29.7±8.2	–	$14$0.70 run	9.82	T0466 · 24 Jan 2026prompt.txt
Business Letters	44.0±6.4	–	$12$0.71 run	4.04	T0103 · 24 Jan 2026prompt.txt
Company Listsolder run	71.0	–	$23$0.34 run	18.18	T0375 · 28 Oct 2025prompt.txt
Fraktur Advertsolder run	54.7	–	–– run	41.93	T0099 · 23 May 2025prompt_optimized.txt
Library Cards	84.3±1.8	–	$10$2.72 run	5.1	T0148 · 24 Jan 2026prompt.txt
Magazine Pages	0.0±0.0	–	$7.79$0.36 run	4.02	T0776 · 23 Mar 2026prompt.txt
Medieval Manuscripts	59.1±16.5	–	$13$0.15 run	11.67	T0291 · 24 Jan 2026prompt.txt
Personnel Cards	92.4±3.6	–	$26$1.60 run	10.33	T0580 · 9 Feb 2026prompt.txt

Not run on: Duty Rosters, General Meeting Minutes.

Price

From	Input, $ per million tokens	Output, $ per million tokens	Source
0 Sep 2025	3.0	15.0	archived price page

Every run