UI-Vision - Functional: leaderboard
Accuracy (%) at pointing inside the target element when the query describes what the element does instead of naming it, 1,772 queries per model.
Metric: Element grounding accuracy (%), Functional setting, 1,772 queries. Source: raw.githubusercontent.com. Saturation forecast: Around 2029. 12 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Aguvis-7B | 18.3 | |
| 2 | CogAgent-9B-20241220 | 12.2 | |
| 3 | OS-Atlas-Base-7B | 11.2 | |
| 4 | Claude 3.7 Sonnet (20250219) | 7.73 | #421 |
| 5 | ShowUI-2B | 7.67 | |
| 6 | MiniCPM-V-2.6 | 5.3 | #3523 |
| 7 | SeeClick | 4.68 | |
| 8 | Qwen 2 VL 7B Instruct | 3.22 | #2594 |
| 9 | InternVL2.5-8B | 2.82 | #2401 |
| 10 | Qwen 2.5 VL 7B Instruct | 0.79 | #1893 |
| 11 | Gemini 2.0 Flash (001) | 0.4 | #727 |
| 12 | InternVL2-8B | 0.11 | #3359 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=ui-vision-functional · How It Works · Data refreshed daily, snapshot 2026-11-17.