Olfactory Perception Benchmark - Olfactory Receptor Activation: leaderboard
Olfactory Perception task: identify which human olfactory receptors a molecule activates among 4 to 10 candidates, from experimental receptor data. Per-question multilabel F1 in percent, higher is better.
Metric: Score (%) on olfactory receptor activation (which of 4 to 10 human receptors a molecule activates, 80 questions from M2OR; per-question multilabel F1) of 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
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.2 Pro | High | 52.8 | #139 |
| 2 | Claude Opus 4.6 | Max | 51.1 | #114 |
| 3 | Claude Opus 4.5 | High | 45.8 | #165 |
| 4 | GPT-5 Pro | 42.5 | #168 | |
| 5 | O3 | High | 42.4 | #242 |
| 6 | Gemini 2.5 Pro | Thinking, 32K | 42.4 | #280 |
| 7 | Grok 3 Mini | High | 41.9 | #559 |
| 8 | GPT-5 | High | 40.8 | #185 |
| 9 | O4 Mini | High | 40.5 | #372 |
| 10 | Claude Sonnet 4.5 | 38.4 | #267 | |
| 11 | GPT-OSS-120B | High | 35.6 | #712 |
| 12 | Llama 3.3 70B Instruct | 35 | #1394 | |
| 13 | DeepSeek Reasoner (Thinking 32K) (Olfactory Perception checkpoint unspecified) | 31.6 | ||
| 14 | Grok 4.1 Fast | 31.1 | #431 | |
| 15 | DeepSeek Reasoner (Thinking 8K) (Olfactory Perception checkpoint unspecified) | 30.5 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=olfactory-perception-benchmark-olfactory-receptor-activation · How It Works · Data refreshed daily, snapshot 2026-11-22.