UI-Vision - Layout Grounding: leaderboard

The model draws a bounding box around a described functional region of a desktop app (a toolbar, a sidebar); the number is mean IoU (%) with the annotated box over the 311 layout queries. Unlike element grounding it asks for whole UI regions, not single controls.

Metric: Layout grounding IoU (%) between predicted and annotated UI-region boxes. Source: raw.githubusercontent.com. Saturation forecast: Estimated already saturated. 7 models tracked. Updated 17 Nov 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.0 Flash (001)28.3#727
2OS-Atlas-Base-7B28.2
3Qwen 2 VL 7B Instruct24.3#2594
4Claude 3.7 Sonnet (20250219)17.6#421
5MiniCPM-V-2.616.3#3523
6CogAgent-9B-202412206.22
7SeeClick5.11

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-layout-grounding · How It Works · Data refreshed daily, snapshot 2026-11-17.