BAGEL - Wikipedia: leaderboard
BAGEL questions on encyclopedic species facts from English Wikipedia articles: taxonomy, behavior, communication, morphology, habitat, cognition, distribution and diet. Four-option closed-book questions written by GPT-4o-mini from source material that is withheld at test time; accuracy in percent, higher is better.
Metric: Accuracy (%, times 100) on the 1,927 Wikipedia questions (encyclopedic species facts in eight dimensions) of BAGEL, closed-book four-option single-answer multiple choice (the source passage is withheld), one unified prompt, greedy decoding with seed 0, a single run per model; questions written by GPT-4o-mini from Wikipedia species articles, GloBI interaction records, bioRxiv preprints and Xeno-canto recordings, options shuffled; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 13 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 4.6 | 93.05 | #113 | |
| 2 | GPT-5.4 | 89.88 | #135 | |
| 3 | Qwen 3 32B | Non-reasoning | 81.06 | #972 |
| 4 | Gemma 3 27B (IT) | 80.12 | #1324 | |
| 5 | Phi-4 | 79.71 | #1931 | |
| 6 | Qwen 3 14B | Non-reasoning | 79.4 | #1269 |
| 7 | Qwen 3 8B | Non-reasoning | 76.54 | #1873 |
| 8 | Llama 3.1 8B Instruct | 73.48 | #4600 | |
| 9 | Mistral 7B Instruct (v0.3) | 68.34 | #6750 | |
| 10 | Qwen 3 4B | Non-reasoning | 67.67 | #2348 |
| 11 | gemma-7B | 64.87 | #7121 | |
| 12 | Qwen 3 0.6B | Non-reasoning | 25.32 | #8350 |
| 13 | SmolLM2-360M | 23.72 | #9570 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=bagel-wikipedia · How It Works · Data refreshed daily, snapshot 2026-11-21.