Released date

Model

Last Updated 4/18/2024

Qwen 2.5 Instruct Turbo, 72B parameters with FP8 quantization.

Released Date: 4/18/2024

Avg. Accuracy:

73.4%

Latency:

2.50s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

79.2%

( 5 / 43 )

79.2%

5 / 43

63.2%

( 31 / 39 )

63.2%

31 / 39

73.6%

( 3 / 47 )

73.6%

3 / 47

77.4%

( 14 / 24 )

77.4%

14 / 24

Academic Benchmarks

Proprietary Benchmarks (contact us to get access)

Input Cost

$1.20 / M Tokens

Output Cost

$1.20 / M Tokens

Input Cost (per char)

$0.26 / M chars

Output Cost (per char)

$0.32 / M chars