ProLLM - OpenBook Q&A — leaderboard
ProLLM benchmark measuring LLM accuracy on open-book question answering with access to reference documents.
Metric: Score (%). Source: www.prollm.ai. Status: saturation imminent. 85 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok 4 | 94.4 |
| 2 | QwQ-32B | 91.9 |
| 3 | Grok 3 Mini | 86.7 |
| 4 | Kimi K2 | 86.3 |
| 5 | Command A | 86.3 |
| 6 | Grok 3 | 85.9 |
| 7 | GPT-4.1 Mini | 85.1 |
| 8 | DeepSeek V3 | 85.1 |
| 9 | Gemini 2.5 Pro | 84.7 |
| 10 | DeepSeek R1 | 84.7 |
| 11 | GPT-4.1 | 82.7 |
| 12 | O1 Preview | 81 |
| 13 | Gemini 2.5 Flash | 80.6 |
| 14 | O1 | 80.6 |
| 15 | Qwen 2.5 72B Instruct | 80.2 |
Interactive version: theaggregate.ai/benchmark?slug=prollm-openbook-q-a · How the rankings work · Data refreshed daily, snapshot 2026-07-22.