AI for Education Visual Reasoning - pattern completion (2d): leaderboard
Metric: Accuracy (%). Source: www.fab-ai.org. 67 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 86.3 |
| 2 | Gemini 3.6 Flash | 83.6 |
| 3 | GPT-5.6 Sol | 82.2 |
| 4 | Gemini 3.5 Flash | 82.2 |
| 5 | Gemini 3.7 Flash | 79.5 |
| 6 | Gemini 3 Flash | 75.3 |
| 7 | GPT-5.5 | 74 |
| 8 | Gemini 3 Pro | 72.6 |
| 9 | Qwen 3.5 397B A17B | 71.2 |
| 10 | Kimi K3 | 71.2 |
| 11 | GPT-5.4 (Medium) | 69.9 |
| 12 | Qwen 3.6 Plus | 67.1 |
| 13 | Kimi K2.5 | 65.8 |
| 14 | GPT-5.2 (Medium) | 65.8 |
| 15 | Qwen 3.6 27B | 63 |
Interactive version: theaggregate.ai/benchmark?slug=ai-for-education-visual-reasoning-pattern-completion-2d · How It Works · Data refreshed daily, snapshot 2026-09-05.