Released date

Model

Last Updated 5/22/2025

Claude Sonnet 4 (Nonthinking)

Anthropic's latest-generation workhorse model, offering a balance of performance and speed.

Released Date: 5/22/2025

Avg. Accuracy:

71.8%

Latency:

51.62s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

FinanceAgent

43.5%

( 5 / 29 )

43.5%

5 / 29

CorpFin

59.9%

( 17 / 40 )

59.9%

17 / 40

CaseLaw

85.2%

( 6 / 63 )

85.2%

6 / 63

ContractLaw

72.4%

( 7 / 69 )

72.4%

7 / 69

TaxEval

73.5%

( 24 / 50 )

73.5%

24 / 50

MortgageTax

74.6%

( 11 / 30 )

74.6%

11 / 30

Math500

90.3%

( 15 / 46 )

90.3%

15 / 46

AIME

38.5%

( 19 / 40 )

38.5%

19 / 40

MGSM

92.9%

( 3 / 44 )

92.9%

3 / 44

LegalBench

81.5%

( 12 / 68 )

81.5%

12 / 68

MedQA

90.3%

( 15 / 48 )

90.3%

15 / 48

GPQA

69.4%

( 13 / 41 )

69.4%

13 / 41

MMLU Pro

79.4%

( 15 / 41 )

79.4%

15 / 41

LiveCodeBench

59.7%

( 14 / 41 )

59.7%

14 / 41

MMMU

72.6%

( 10 / 27 )

72.6%

10 / 27

SWE-bench

65.0%

( 1 / 12 )

65.0%

1 / 12

Academic Benchmarks

Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$3.00 / M Tokens

Output Cost

$15.00 / M Tokens

Input Cost (per char)

N/A

Output Cost (per char)

$5.23 / M chars

Performance by Benchmark

Cost Analysis

Join our mailing list to receive benchmark updates on