Skip to content

gpt-5.3-codex

OpenAI

average±1.9 (95%): 81.7
leaderboard rankof 114: ≈25
benchmarksof 12: 12
per 1,000 documentsmean over its benchmarks: $13
seconds per inputmean response time: 9.7

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

By task

Data correction

67.2

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

90.4

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

84.0

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

89.5

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	98.2±1.3	≈8of 115	$25$0.12 run	17.16	T0693 · 23 Mar 2026prompt.txt
Blacklist Cards	87.8±3.1	≈79of 114	$5.61$0.19 run	3.14	T0700 · 23 Mar 2026prompt.txt
Book Advert XML Repair	37.6±8.7	≈41of 119	$9.30$0.47 run	7.94	T0703 · 23 Mar 2026prompt.txt
Business Letters	60.0±7.0	≈24of 114	$7.12$0.41 run	3.12	T0694 · 23 Mar 2026prompt.txt
Company Lists	82.7±10.9	≈29of 114	$16$0.24 run	10.98	T0701 · 17 Mar 2026prompt.txt
Duty Rosters	88.4±3.0	≈18of 107	$22$0.46 run	8.07	T1312 · 16 Sep 2026prompt.txt
Fraktur Adverts	93.6±3.6	≈32of 114	$25$0.13 run	27.06	T0697 · 23 Mar 2026prompt_optimized.txt
General Meeting Minutes	85.8±8.6	≈25of 108	$13$0.11 run	15.17	T0705 · 17 Mar 2026prompt.txt
Library Cards	84.9±1.5	≈51of 114	$5.55$1.46 run	2.65	T0698 · 23 Mar 2026prompt.txt
Magazine Pages	86.0±9.0	≈18of 115	$4.15$0.19 run	3.9	T0710 · 21 Mar 2026prompt.txt
Medieval Manuscripts	78.3±8.2	≈27of 114	$7.68$0.092 run	7.36	T0699 · 23 Mar 2026prompt.txt
Personnel Cards	96.7±1.0	≈30of 114	$18$1.10 run	10.08	T0704 · 23 Mar 2026prompt.txt

Price

From	Input, $ per million tokens	Output, $ per million tokens	Source
16 Mar 2026	1.75	14	archived price page

Every run