UI-Venus-2 Report - UI-Vision: leaderboard
UI-Vision evaluates GUI grounding across visually diverse and challenging desktop interface layouts. From Ant Group's UI-Venus-2 technical report, keeping only runs the authors performed or reproduced, with UI-Venus-2 run with reasoning off; accuracy in percent, higher is better.
Metric: Accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.6 27B | 58.3 | #182 |
| 2 | Kimi K2.6 | 51.7 | #113 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ui-venus-2-report-ui-vision · How It Works · Data refreshed daily, snapshot 2026-11-05.