AutoGUI-v2 - Region Captioning: leaderboard

AutoGUI-v2 (2026) tests functionality understanding of GUI regions and elements on web, desktop and mobile screenshots. This board asks which of several options states the function of a region, with easy and hard distractors. Accuracy in percent; higher is better.

Metric: Region functionality captioning on 447 tasks, multiple-choice accuracy (%) with easy and hard (same-group) distractors; the authors' tuned prompts; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-588.1#178
2GLM-4.5V86.8#754
3Gemini 2.5 Pro86.2#276
4Claude Sonnet 4.584#261
5O384#237
6Qwen 3 VL 32B Instruct80.1#606
7Qwen 3 VL 8B Instruct66.3#1029

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=autogui-v2-region-captioning · How It Works · Data refreshed daily, snapshot 2026-11-21.