HELM NaturalQuestions (Open): leaderboard
Open-book question answering from Google search queries with access to Wikipedia passages, scored by F1. Part of Stanford HELM.
Metric: F1 (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 91 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Nova Pro | 82.9 | #595 |
| 2 | Nova Lite | 81.52 | #742 |
| 3 | GPT-4o (2024-05-13) | 80.32 | #337 |
| 4 | GPT-4 Turbo | 79.52 | #418 |
| 5 | GPT-4o (2024-08-06) | 79.26 | #343 |
| 6 | GPT-4 (0613) | 78.97 | #495 |
| 7 | Nova Micro | 77.86 | #891 |
| 8 | Qwen 1.5 32B | 77.71 | #793 |
| 9 | Command | 77.69 | #1297 |
| 10 | Qwen 2 72B Instruct | 77.58 | #580 |
| 11 | Yi 34B (Base) | 77.51 | #952 |
| 12 | Qwen 1.5 14B | 77.21 | #1082 |
| 13 | text-davinci-003 | 76.98 | #1107 |
| 14 | DeepSeek V3 | 76.53 | #333 |
| 15 | GPT-4 Turbo (Preview) | 76.29 | #417 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-naturalquestions-open · How It Works · Data refreshed daily, snapshot 2026-11-06.