HELM Classic - QuAC — leaderboard

Metric: F1 (%). Source: crfm.stanford.edu. 66 models tracked.

Top models

#ModelScore
1text-davinci-00352.48
2GPT-3.5 Turbo (0301)51.2
3GPT-3.5 Turbo (0613)48.5
4Llama 2 70B48.44
5text-davinci-00244.51
6Llama 2 13B42.38
7Mistral-7B-v0.142.32
8Llama 2 7B40.62
9LLaMA-65B40.13
10mpt-30B39.31
11LLaMA-30B39.02
12davinci36.04
13text-curie-00135.79
14LLaMA-13B34.75
15LLaMA-7B33.76

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