Released date

Model

Last Updated 5/22/2025

Claude Sonnet 4

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

Released Date: 5/22/2025

Avg. Accuracy:

76.9%

Latency:

9.92s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

CaseLaw

85.2%

( 5 / 60 )

85.2%

5 / 60

ContractLaw

72.4%

( 8 / 67 )

72.4%

8 / 67

TaxEval

73.5%

( 22 / 47 )

73.5%

22 / 47

Math500

90.3%

( 12 / 43 )

90.3%

12 / 43

AIME

38.5%

( 16 / 37 )

38.5%

16 / 37

MGSM

92.9%

( 2 / 41 )

92.9%

2 / 41

LegalBench

81.5%

( 10 / 65 )

81.5%

10 / 65

MedQA

90.3%

( 12 / 45 )

90.3%

12 / 45

GPQA

69.4%

( 10 / 38 )

69.4%

10 / 38

MMLU Pro

79.4%

( 12 / 38 )

79.4%

12 / 38

MMMU

72.6%

( 7 / 24 )

72.6%

7 / 24

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)

$0.82 / M chars

Output Cost (per char)

$5.68 / M chars

Performance by Benchmark

Cost Analysis

Join our mailing list to receive benchmark updates on