Olfactory Perception Benchmark - Odor Primary Descriptor: leaderboard
Olfactory Perception task: choose the primary odor descriptor of a molecule from four options, using the IFRA fragrance ingredient glossary. Accuracy in percent, higher is better.
Metric: Score (%) on odor primary descriptor identification (one of four options, 175 molecules from the IFRA glossary; any-overlap accuracy) 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 February 2027. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 2.5 Pro | Thinking, 8K | 80 | #276 |
| 2 | Claude Opus 4.6 | High | 78.3 | #113 |
| 3 | Claude Opus 4.5 | High | 76.6 | #160 |
| 4 | GPT-5.2 Pro | High | 76 | #137 |
| 5 | GPT-5 | High | 73.7 | #178 |
| 6 | DeepSeek Reasoner (Thinking 32K) (Olfactory Perception checkpoint unspecified) | 73.7 | ||
| 7 | GPT-5 Pro | 73.1 | #163 | |
| 8 | Grok 3 Mini | Low | 73.1 | #565 |
| 9 | O3 | High | 70.9 | #237 |
| 10 | DeepSeek Reasoner (Thinking 8K) (Olfactory Perception checkpoint unspecified) | 70.3 | ||
| 11 | DeepSeek Reasoner (Thinking 16K) (Olfactory Perception checkpoint unspecified) | 69.7 | ||
| 12 | Claude Sonnet 4.5 | 67.4 | #261 | |
| 13 | Grok 4.1 Fast | 67.4 | #427 | |
| 14 | O4 Mini | High | 65.7 | #368 |
| 15 | Llama 3.3 70B Instruct | 60.6 | #1385 |
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-odor-primary-descriptor · How It Works · Data refreshed daily, snapshot 2026-11-21.