CVSBench - FOV Satellite-to-Ground: leaderboard

Metric: Accuracy (%) on FOV-subset satellite-to-ground questions (limited field-of-view street images); four-choice and binary-choice questions over paired street-view and satellite images (Gemini-2.5-flash-generated, human-verified), test split; higher is better. Source: arxiv.org. Saturation forecast: Around 2032. 15 models tracked.

Top models

#ModelScore
1GPT-5 Chat54.1
2Gemma 3 4B (IT)47.99
3GPT-4o46.8
4Qwen 3 VL 8B Instruct46.7
5Qwen 3 VL 4B Instruct44.2
6InternVL3.5-8B41.5
7DeepSeek V3.238.68

Interactive version: theaggregate.ai/benchmark?slug=cvsbench-fov-satellite-to-ground · How It Works · Data refreshed daily, snapshot 2026-09-29.