New Finance Agent Benchmark Released
openai/gpt-5-mini-2025-08-07 GPT 5 Mini

Faster and a more cost-efficient version of GPT-5.

Released Date: 8/7/2025

Avg. Accuracy:

80.9%

Latency:

144.19s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

CaseLaw

77.5%

( 2 / 38 )

TaxEval

80.1%

( 1 / 72 )

MortgageTax

79.2%

( 2 / 44 )

Math500

94.8%

( 5 / 54 )

AIME

90.8%

( 4 / 62 )

MGSM

92.6%

( 10 / 65 )

LegalBench

81.7%

( 16 / 87 )

MedQA

96.2%

( 3 / 68 )

GPQA

80.3%

( 7 / 64 )

MMLU Pro

82.5%

( 16 / 62 )

MMMU

78.9%

( 8 / 41 )

LiveCodeBench

86.6%

( 1 / 61 )

Terminal-Bench

30.0%

( 13 / 23 )

Academic Benchmarks
Proprietary Benchmarks (contact us to get access)

Cost Analysis

Input Cost

$0.25 / M Tokens

Output Cost

$2.00 / M Tokens

Join our mailing list to receive benchmark updates on

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