claude-3-5-sonnet-20241022

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: 68.8
benchmarksof 12: 7
per 1,000 documentsmean over its benchmarks: $20
seconds per inputmean response time: 12.6

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

By benchmark

Benchmark	Score	Rank	Per 1,000 docs	s / input	Best run
Bibliographic Dataolder run	94.9	–	$34$0.17 run	24.84	T0009 · 30 Sep 2025prompt.txt
Blacklist Cardsolder run	65.8	–	$8.76$0.29 run	4.3	T0319 · 24 Oct 2025prompt.txt
Business Lettersolder run	45.0	–	–– run	3.18	T0018 · 5 Mar 2025prompt.txt
Company Listsolder run	66.6	–	$22$0.33 run	15.59	T0367 · 28 Oct 2025prompt.txt
Fraktur Advertsolder run	49.8	–	$26$0.13 run	26.61	T0093 · 30 Sep 2025prompt_optimized.txt
Library Cardsolder run	88.5	–	–– run	5.57	T0143 · 2 Sep 2025prompt.txt
Medieval Manuscriptsolder run	71.0	–	$10$0.12 run	8.06	T0288 · 24 Oct 2025prompt.txt

Not run on: Book Advert XML Repair, Duty Rosters, General Meeting Minutes, Magazine Pages, Personnel Cards.

Price

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

claude-3-5-sonnet-20241022

By benchmark

Price

Every run