0

Qwen2.5 72B Instruct

Qwen2.5 72B Instruct is an AI model from Alibaba.

Publisher
Alibaba
Family
Qwen2.5
Input
Output
License
Qwen
Released
19 Sep 2024

Cite

Notes

Only stored in your browser.

Attribution

Benchmark scores
AAarena-hard-autoOpenLLM
Attribution policy →
Weights
Closed
API only
Input
$0.47
/1M tok
Output
$0.49
/1M tok
Blended
$0.48
#334/ 499/1M · 3:1
Speed
0
#103/ 498tok/s

Price & speed via Artificial Analysis · ranked across tracked models

Reported on 16 evals across 7 domains - best 86.4% on IFEval · #3 of 81

Reported eval scores

16
BIG-Bench Hard (BBH)72.7
IFEval86.4
MuSR42.1
MMLU-Pro56.3
Humanity's Last Exam (HLE)3.6
Arena-Hard10.2
MATH Level 559.8
Terminal-Bench (Hard)4.5
IFBench36.9
MATH-50085.8
AIME 2024: Problems from the American Invitational Mathematics Examination16
τ²-bench (Tau²-bench)34.5
LiveCodeBench27.6
SciCode26.7
GPQA Diamond37.5
AIME 2025: Problems from the American Invitational Mathematics Examination14