New Finance Agent Benchmark Released

Llama 3.3 Instruct Turbo, 70B parameters with FP16 quantization.

Released Date: 12/6/2024

Avg. Accuracy:

56.6%

Latency:

6.83s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

FinanceAgent

3.4%

( 24 / 26 )

CaseLaw

76.6%

( 39 / 62 )

ContractLaw

66.0%

( 37 / 68 )

TaxEval

63.9%

( 39 / 49 )

Math500

73.4%

( 35 / 46 )

AIME

16.0%

( 28 / 40 )

MGSM

91.3%

( 17 / 43 )

LegalBench

78.0%

( 28 / 67 )

GPQA

50.0%

( 28 / 41 )

MMLU Pro

69.9%

( 27 / 40 )

LiveCodeBench

34.3%

( 34 / 41 )

Academic Benchmarks
Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$0.88 / M Tokens

Output Cost

$0.88 / M Tokens

Input Cost (per char)

$0.49 / M chars

Output Cost (per char)

$0.23 / M chars

Join our mailing list to receive benchmark updates on

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