BAGEL - Wikipedia Morphology: leaderboard

BAGEL Wikipedia questions on body form and appearance of a species. 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 233 Wikipedia morphology questions 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: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Opus 4.689.3#113
2GPT-5.486.3#135
3Qwen 3 14BNon-reasoning73.4#1269
4Qwen 3 32BNon-reasoning71.2#972
5Gemma 3 27B (IT)70.8#1324
6Phi-470.4#1931
7Qwen 3 8BNon-reasoning69.1#1873
8Llama 3.1 8B Instruct66.5#4600
9Qwen 3 4BNon-reasoning60.5#2348
10gemma-7B54.5#7121
11Mistral 7B Instruct (v0.3)54.1#6750
12Qwen 3 0.6BNon-reasoning29.2#8350
13SmolLM2-360M27.9#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-morphology · How It Works · Data refreshed daily, snapshot 2026-11-21.