HELM Classic - QuAC: leaderboard
The "QuAC" subtask of HELM Classic. A scenario from Stanford HELM Classic, the holistic evaluation suite scoring language models across accuracy, robustness and calibration on a broad set of tasks.
Metric: F1 (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 66 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | text-davinci-003 | 52.48 | #1107 |
| 2 | GPT-3.5 Turbo (0613) | 48.5 | #1113 |
| 3 | Llama 2 70B | 48.44 | #1044 |
| 4 | Palmyra X (43B) | 47.29 | #1255 |
| 5 | text-davinci-002 | 44.51 | #1200 |
| 6 | Llama 2 13B | 42.38 | #1414 |
| 7 | Mistral-7B-v0.1 | 42.32 | #1295 |
| 8 | Llama 2 7B | 40.62 | #1499 |
| 9 | vicuna-13B-v1.3 | 40.27 | #1429 |
| 10 | LLaMA-65B | 40.13 | #1266 |
| 11 | mpt-30B | 39.31 | #1454 |
| 12 | vicuna-7B-v1.3 | 39.22 | #1485 |
| 13 | LLaMA-30B | 39.02 | #1397 |
| 14 | falcon-40B Instruct | 37.07 | #1441 |
| 15 | davinci | 36.04 | #1533 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-classic-quac · How It Works · Data refreshed daily, snapshot 2026-11-06.