iOSWorld (Vision+XML) - Single-App: leaderboard

Metric: Task pass rate (%) on the 27 single-app tasks, vision+XML; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 6 models tracked.

Top models

#ModelScore
1Claude Sonnet 4.692.6
2GPT-5.481.5
3Claude Opus 4.681.5
4Gemini 3 Flash70.4
5GPT-5.4 Mini66.7
6Qwen 3.5 35B A3B48.1

Interactive version: theaggregate.ai/benchmark?slug=iosworld-vision-plusxml-single-app · How It Works · Data refreshed daily, snapshot 2026-09-29.