SkateBench: leaderboard
Skateboarding-domain knowledge benchmark where models identify technical skateboard tricks from 390 concise definitions, including subtle differences in stance, rotation, flips, and grinds.
Metric: Success Rate (%). Source: skatebench.t3.gg. Saturation forecast: Around January 2027. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 96.92 | #58 |
| 2 | GPT-5.4 (High) | 81.54 | #79 (GPT-5.4) |
| 3 | GPT-5.4 (xHigh) | 81.28 | #79 (GPT-5.4) |
| 4 | GPT-5.4 (Medium) | 78.46 | #79 (GPT-5.4) |
| 5 | GPT-5 (High) | 77.44 | #98 (GPT-5) |
| 6 | Gemini 3 Pro (Preview) | 76.15 | #66 |
| 7 | GPT-5 | 75.38 | #98 |
| 8 | Gemini 3 Flash (High) | 75.13 | #99 (Gemini 3 Flash) |
| 9 | GPT-5.1 (High) | 71.79 | #135 (GPT-5.1) |
| 10 | GPT-5.2 Pro | 71.79 | #61 |
| 11 | GPT-5.2 (xHigh) | 68.97 | #110 (GPT-5.2) |
| 12 | GLM-5 | 66.67 | #141 |
| 13 | Claude Opus 4.6 (Thinking, High) | 64.36 | #63 (Claude Opus 4.6) |
| 14 | Grok 4 | 61.28 | #181 |
| 15 | Kimi K2.5 | 57.18 | #144 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=skatebench · How It Works · Data refreshed daily, snapshot 2026-10-20.