Skip to content

gpt-5-nano

OpenAI

average: 53.5
leaderboard rankof 114: ≈101
benchmarksof 12: 12
per 1,000 documentsmean over its benchmarks: $2.35
seconds per inputmean response time: 41.7

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

By task

Data correction

47.2

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

57.6

Best of 107 models with all its benchmarks: gpt-6-astra, 97.0.
Information extraction

67.4

Best of 107 models with all its benchmarks: grok-4.6, 90.8.
Transcription

35.8

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 Data	90.1±5.7	≈74of 115	$5.27$0.026 run	70.67	T0131 · 26 Sep 2026prompt.txt
Blacklist Cards	76.2±4.1	≈106of 114	$1.19$0.039 run	29.27	T0311 · 26 Jan 2026prompt.txt
Book Advert XML Repair	13.1±6.6	≈89of 119	$4.84$0.24 run	108.4	T0453 · 8 Dec 2025prompt.txt
Business Letters	52.0±7.2	≈95of 114	$1.22$0.069 run	20.61	T0114 · 23 Mar 2026prompt.txt
Company Lists	71.4±13.6	≈84of 114	$2.61$0.039 run	52.69	T0351 · 25 Jan 2026prompt.txt
Duty Rosters	41.5±12.7	≈96of 107	$1.96$0.041 run	35.95	T1300 · 24 Jul 2026prompt.txt
Fraktur Adverts	1.6±3.2	≈104of 114	$1.16$0.006 run	23.68	T0122 · 25 Jan 2026prompt_optimized.txt
General Meeting Minutes	60.6±18.4	≈85of 108	$3.05$0.027 run	58.95	T1251 · 3 Jul 2026prompt.txt
Library Cardsolder run	79.7	≈95of 114	–– run	18.98	T0167 · 2 Sep 2025prompt.txt
Magazine Pages	50.2±12.0	≈47of 115	$0.84$0.039 run	13.87	T0777 · 23 Mar 2026prompt.txt
Medieval Manuscripts	24.4±14.8	≈107of 114	$0.88$0.011 run	22.6	T0281 · 26 Jan 2026prompt.txt
Personnel Cards	81.2±8.5	≈94of 114	$2.82$0.17 run	44.8	T0612 · 16 Sep 2026prompt.txt

Price

From	Input, $ per million tokens	Output, $ per million tokens	Source
0 Sep 2025	0.05	0.4	archived price page

Every run