BoolQ: leaderboard
Boolean Questions: 15,942 naturally occurring yes/no questions from Google search queries paired with Wikipedia passages. Tests reading comprehension and inference.
Metric: Accuracy (%). Source: huggingface.co. Source: epoch.ai. Saturation forecast: Estimated already saturated. 77 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | T5 (11B) | 91.2 | #1652 |
| 2 | Median Human | 90 | |
| 3 | StableBeluga2 | 89.4 | #1078 |
| 4 | GPT-4o Mini (2024-07-18) | 88.7 | #558 |
| 5 | Llama 2 70B Base | 88.6 | #1025 |
| 6 | text-davinci-003 | 88.1 | #1146 |
| 7 | text-davinci-002 | 87.7 | #1225 |
| 8 | internlm-20B | 87.5 | #1296 |
| 9 | Mistral-7B-v0.1 | 87.4 | #1326 |
| 10 | LLaMA-65B | 87.1 | #1303 |
| 11 | GPT-3.5 Turbo (0613) | 87 | #1118 |
| 12 | Qwen-14B | 86.2 | #1171 |
| 13 | Gemini 1.5 Flash (001) | 85.8 | #577 |
| 14 | Gemma 2 9B | 85.7 | #958 |
| 15 | mpt-30B Instruct | 85 | #1470 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=boolq · How It Works · Data refreshed daily, snapshot 2026-10-20.