New Finance Agent Benchmark Released

Released date

Model

Last Updated 5/22/2025

anthropic/claude-opus-4-20250514

Claude Opus 4 (Nonthinking)

Anthropic's most powerful model.

Released Date: 5/22/2025

Avg. Accuracy:

73.8%

Latency:

20.95s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

42.3%

( 7 / 29 )

42.3%

7 / 29

81.5%

( 22 / 63 )

81.5%

22 / 63

68.8%

( 15 / 69 )

68.8%

15 / 69

75.8%

( 13 / 50 )

75.8%

13 / 50

70.9%

( 18 / 30 )

70.9%

18 / 30

90.4%

( 14 / 46 )

90.4%

14 / 46

41.3%

( 17 / 40 )

41.3%

17 / 40

93.8%

( 1 / 44 )

93.8%

1 / 44

81.7%

( 10 / 68 )

81.7%

10 / 68

92.9%

( 7 / 48 )

92.9%

7 / 48

71.7%

( 12 / 41 )

71.7%

12 / 41

86.1%

( 1 / 41 )

86.1%

1 / 41

62.6%

( 11 / 41 )

62.6%

11 / 41

73.4%

( 8 / 27 )

73.4%

8 / 27

Academic Benchmarks

Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$15.00 / M Tokens

Output Cost

$75.00 / M Tokens

Input Cost (per char)

N/A

Output Cost (per char)

$27.05 / M chars

Join our mailing list to receive benchmark updates on

Stay up to date as new benchmarks and models are released.