BAGEL - bioRxiv: leaderboard

BAGEL questions asking for the interpretation of results in animal-focused bioRxiv preprints from behavior, ecology, evolution and zoology. 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 2,183 bioRxiv questions (result interpretation of animal-focused preprints) 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

#ModelSettingScoreOverall rank
1GPT-5.494.41#138
2Phi-492.44#1951
3Qwen 3 32BNon-reasoning91.85#983
4Qwen 3 14BNon-reasoning91.11#1284
5Claude Opus 4.691.07#114
6Gemma 3 27B (IT)90.98#1341
7Qwen 3 8BNon-reasoning89.92#1890
8Qwen 3 4BNon-reasoning89.01#2379
9Llama 3.1 8B Instruct88.82#4655
10Mistral 7B Instruct (v0.3)85.57#6759
11gemma-7B85.43#7115
12SmolLM2-360M25.1#9572
13Qwen 3 0.6BNon-reasoning24.05#8357

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=bagel-biorxiv · How It Works · Data refreshed daily, snapshot 2026-11-22.