Olfactory Perception Benchmark: leaderboard

The Olfactory Perception benchmark (2026) tests whether language models can reason about smell from molecule names: odor detectability, descriptors, intensity and pleasantness, descriptor profiles, mixture similarity, receptor activation and smell identification. The score is the unweighted mean of the eight task scores in percent, higher is better.

Metric: Overall score (%): unweighted mean of the eight task scores (accuracy for single-answer tasks, multilabel F1 for rate-all-that-apply and receptor activation; the odor similarity of mixtures task, where no model exceeds chance on similar pairs, is included in the mean) over the 1,010 questions of the Olfactory Perception (OP) benchmark presented with compound-name prompts (the isomeric SMILES variant is reported only in figures), constrained option lists, provider APIs without web search or tools, each model at its stated reasoning effort or thinking budget; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 16 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Opus 4.6Max64.4#113
2GPT-5.2 ProHigh62.2#137
3Claude Opus 4.5High62#160
4GPT-5 Pro61.9#163
5GPT-5High61.1#178
6Gemini 2.5 ProThinking, 32K59.7#276
7Claude Sonnet 4.559.6#261
8O3High59.3#237
9O4 MiniHigh59#368
10DeepSeek Reasoner (Thinking 32K) (Olfactory Perception checkpoint unspecified)58.5
11Grok 4.1 Fast58.3#427
12Grok 3 MiniHigh57.9#565
13DeepSeek Reasoner (Thinking 8K) (Olfactory Perception checkpoint unspecified)57.3
14DeepSeek Reasoner (Thinking 16K) (Olfactory Perception checkpoint unspecified)56.6
15GPT-OSS-120BHigh54#704

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=olfactory-perception-benchmark · How It Works · Data refreshed daily, snapshot 2026-11-21.