OpenBookQA (PapersWithCode): leaderboard
OpenBookQA (Mihaylov et al., 2018): four-option elementary science questions that need a provided book of elementary science facts plus broader common knowledge. This board is the archived Papers With Code table of accuracies as each paper reported them, under that paper’s own setup (some rows add external knowledge), not a common evaluation run; higher is better.
Metric: Accuracy. Source: huggingface.co. Saturation forecast: Rough model projection: around 2026. 45 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 | 95.9 | #459 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openbookqa-paperswithcode · How It Works · Data refreshed daily, snapshot 2026-11-05.