BuzzBench (Humour): leaderboard
Humour analysis benchmark testing how well LLMs can understand and analyze comedy - a component of emotional intelligence.
Metric: Humour Score (0-100). Source: eqbench.com. Saturation forecast: Estimated already saturated. 26 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 71.09 | #123 |
| 2 | Claude 3.7 Sonnet (20250219) | 68.2 | #201 |
| 3 | GPT-4.1 | 67.1 | #248 |
| 4 | DeepSeek R1 | 62.73 | #251 |
| 5 | Claude 3.5 Sonnet | 61.94 | #350 |
| 6 | GPT-4o ChatGPT | 59.48 | |
| 7 | O1 | 59.28 | #183 |
| 8 | Gemini 1.5 Pro | 54.9 | #445 |
| 9 | Gemini 2.0 Flash (001) | 54.39 | #360 |
| 10 | Claude 3.5 Haiku (20241022) | 53.34 | #588 |
| 11 | Grok 2 (1212) | 51.76 | #496 |
| 12 | DeepSeek V3 | 51.63 | #321 |
| 13 | Mistral Large 2 (Nov) Instruct (2411) | 47.48 | #465 |
| 14 | O1 Mini | 46.18 | #323 |
| 15 | Llama 3.1 405B Instruct | 45.94 | #457 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=buzzbench-humour · How It Works · Data refreshed daily, snapshot 2026-10-20.