CVSBench - FOV Ground-to-Satellite: leaderboard
CVSBench (2026) tests cross-view spatial reasoning with paired street-view and satellite images: models infer attributes of the unseen view and localize the street-view camera on the satellite image, on four-choice and binary questions. This board is accuracy in percent on FOV-subset questions (limited field-of-view street images) that infer satellite-view attributes; higher is better.
Metric: Accuracy (%) on FOV-subset ground-to-satellite 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 April 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.2 | 56.69 | #406 |
| 2 | GPT-5 Chat | 53.5 | #444 |
| 3 | Gemma 3 4B (IT) | 40.51 | #3831 |
| 4 | GPT-4o | 36 | #741 |
| 5 | Qwen 3 VL 4B Instruct | 26.6 | #1279 |
| 6 | Qwen 3 VL 8B Instruct | 24 | #981 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=cvsbench-fov-ground-to-satellite · How It Works · Data refreshed daily, snapshot 2026-11-13.