MobileJudgeBench (A3 Final-State Judge): leaderboard

Metric: Judge accuracy (%; share of 931 human-annotated mobile-agent trajectories, from 6 benchmarks, 4 agent models and 68 apps, whose success or failure the judge classifies correctly; the A3 judge in final-state mode, temperature 0.7 as in its original implementation). Source: arxiv.org. Saturation forecast: Around December 2026. 5 models tracked.

Top models

#ModelScore
1Gemini 3 Flash (Preview)83.8
2Qwen 2.5 VL 72B Instruct81.5
3Claude Sonnet 4.578.4
4GLM-4.6V78.1
5GPT-5 Mini76.4

Interactive version: theaggregate.ai/benchmark?slug=mobilejudgebench-a3-final-state-judge · How It Works · Data refreshed daily, snapshot 2026-09-26.