UI-Venus-2 Report - VenusBench-Mobile: leaderboard
VenusBench-Mobile evaluates general-purpose mobile GUI agents on realistic, user-centric everyday phone tasks; success rate on its 149-task primary pool. From Ant Group's UI-Venus-2 technical report, keeping only UI-Venus-2's own runs and baselines the authors evaluated or reproduced; percent, higher is better.
Metric: Success Rate (%). Source: arxiv.org. Saturation forecast: Around December 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 36.5 | #67 |
| 2 | Kimi K2.6 | 31.2 | #113 |
| 3 | Qwen 3.6 27B | 28 | #182 |
| 4 | Seed 2.0 Pro | 20.1 | #122 |
| 5 | Qwen 3.5 9B | 15.3 | #391 |
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-venusbench-mobile · How It Works · Data refreshed daily, snapshot 2026-11-05.