BlueBench - News Classification — leaderboard

Metric: Score (%). Source: huggingface.co. 18 models tracked.

Top models

#ModelScore
1GPT-4o69.07
2O168.16
3O4 Mini67.37
4GPT-4.1 Mini66.67
5GPT-4.166.29
6Llama 3.3 70B Instruct59.89
7Mistral Medium 353.76
8O3 Mini43.06
9GPT-4.1 Nano42.16
10Llama 3.2 3B Instruct36.47
11Mistral Large29.01
12Pixtral-12B21.54
13Llama 3.2 1B Instruct20

Interactive version: theaggregate.ai/benchmark?slug=bluebench-news-classification · How the rankings work · Data refreshed daily, snapshot 2026-07-22.