VG-GUI-Bench (No Video): leaderboard

Metric: Step accuracy (%; 0.3 for the right action type plus up to 0.7 for its arguments, over the 1,071 steps of 100 mobile GUI episodes; no tutorial frames, current screen only). Source: arxiv.org. Saturation forecast: Around March 2027. 7 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)58.51
2Kimi K2.557.22
3GPT-5 Mini55.76
4Claude Sonnet 4.644.35
5Seed 2.0 Pro35.93
6Qwen 3 VL 235B A22B25.9

Interactive version: theaggregate.ai/benchmark?slug=vg-gui-bench-no-video · How It Works · Data refreshed daily, snapshot 2026-09-26.