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: epoch.ai. Status: saturated. 77 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Median Human | 90 |
| 2 | StableBeluga2 | 89.4 |
| 3 | falcon-180B | 89 |
| 4 | GPT-4o Mini (2024-07-18) | 88.7 |
| 5 | Llama 2 70B Base | 88.6 |
| 6 | text-davinci-003 | 88.1 |
| 7 | text-davinci-002 | 87.7 |
| 8 | Mistral-7B-v0.1 | 87.4 |
| 9 | LLaMA-65B | 87.1 |
| 10 | GPT-3.5 Turbo (0613) | 87 |
| 11 | Qwen-14B | 86.2 |
| 12 | Gemini 1.5 Flash (001) | 85.8 |
| 13 | Gemma 2 9B | 85.7 |
| 14 | mpt-30B Instruct | 85 |
| 15 | Phi-3.5-MoE-instruct | 84.6 |
Interactive version: theaggregate.ai/benchmark?slug=boolq · How the rankings work · Data refreshed daily, snapshot 2026-07-22.