HELM Classic - RAFT — leaderboard

Metric: Exact Match (%). Source: crfm.stanford.edu. 67 models tracked.

Top models

#ModelScore
1GPT-3.5 Turbo (0301)76.82
2text-davinci-00375.91
3LLaMA-30B75.23
4GPT-3.5 Turbo (0613)74.77
5text-davinci-00273.26
6Llama 2 70B72.73
7mpt-30B72.27
8Mistral-7B-v0.170.68
9Llama 2 13B70.68
10LLaMA-65B70.23
11falcon-40B66.14
12LLaMA-13B64.32
13Llama 2 7B64.32
14davinci64.24
15gpt-j-6B61.89

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