RABBITS B4BQA: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 26 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4 | 99.71 |
| 2 | GPT-4o | 99.61 |
| 3 | GPT-3.5 Turbo | 99.43 |
| 4 | Claude 3 Opus | 99.14 |
| 5 | Qwen 2 72B | 99.05 |
| 6 | Mixtral-8x22B-v0.1 | 98.66 |
| 7 | Llama 3 70B | 98.57 |
| 8 | Command-R+ | 98.28 |
| 9 | c4ai-command-r-plus | 98.28 |
| 10 | Yi 1.5 34B | 97.81 |
| 11 | Mixtral 8x7B (v0.1) | 97.14 |
| 12 | Llama 3 8B | 96.18 |
| 13 | aya-23-35B | 95.81 |
| 14 | Llama 2 70B Base | 93.61 |
| 15 | Qwen 2 7B | 93.13 |
Interactive version: theaggregate.ai/benchmark?slug=rabbits-b4bqa · How It Works · Data refreshed daily, snapshot 2026-09-05.