New Finance Agent Benchmark Released

Released date

Model

Last Updated 9/23/2025

alibaba/qwen3-max

Largest and most recent Qwen Model

Released Date: 9/23/2025

Avg. Accuracy:

63.0%

Latency:

407.62s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

32.7%

( 14 / 39 )

32.7%

14 / 39

62.8%

( 22 / 52 )

62.8%

22 / 52

66.6%

( 22 / 35 )

66.6%

22 / 35

76.7%

( 15 / 69 )

76.7%

15 / 69

77.9%

( 16 / 59 )

77.9%

16 / 59

91.8%

( 18 / 62 )

91.8%

18 / 62

38.1%

( 82 / 84 )

38.1%

82 / 84

87.4%

( 34 / 65 )

87.4%

34 / 65

79.5%

( 8 / 61 )

79.5%

8 / 61

83.7%

( 9 / 59 )

83.7%

9 / 59

70.4%

( 15 / 58 )

70.4%

15 / 58

15.7%

( 4 / 18 )

15.7%

4 / 18

36.3%

( 10 / 20 )

36.3%

10 / 20

62.4%

( 4 / 21 )

62.4%

4 / 21

Academic Benchmarks

Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$1.20 / M Tokens

Output Cost

$6.00 / M Tokens

Input Cost (per char)

N/A

Output Cost (per char)

$2.44 / M chars

Join our mailing list to receive benchmark updates on

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