Olfactory Perception Benchmark - Odor Intensity: leaderboard
Olfactory Perception task: decide which of two molecules smells more intense, based on human psychophysical ratings. Accuracy in percent, higher is better.
Metric: Score (%) on odor intensity paired comparison (which of two molecules smells more intense, 175 pairs from Keller et al. 2017 ratings; 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 September 2028. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 4.6 | Max | 74.9 | #113 |
| 2 | GPT-5.2 Pro | High | 72.6 | #137 |
| 3 | GPT-5 Pro | 71.4 | #163 | |
| 4 | Claude Opus 4.5 | High | 71.4 | #160 |
| 5 | DeepSeek Reasoner (Thinking 8K) (Olfactory Perception checkpoint unspecified) | 71.4 | ||
| 6 | GPT-5 | Low | 70.9 | #178 |
| 7 | DeepSeek Reasoner (Thinking 32K) (Olfactory Perception checkpoint unspecified) | 69.7 | ||
| 8 | O4 Mini | High | 69.1 | #368 |
| 9 | DeepSeek Reasoner (Thinking 16K) (Olfactory Perception checkpoint unspecified) | 68.6 | ||
| 10 | Llama 3.3 70B Instruct | 68 | #1385 | |
| 11 | O3 | High | 68 | #237 |
| 12 | Grok 3 Mini | High | 68 | #565 |
| 13 | Gemini 2.5 Pro | Thinking, 16K | 68 | #276 |
| 14 | Claude Sonnet 4.5 | 66.9 | #261 | |
| 15 | Grok 4.1 Fast | 66.9 | #427 |
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-intensity · How It Works · Data refreshed daily, snapshot 2026-11-21.