UltraVR - Remote Sensing: leaderboard

Metric: Accuracy (%) on the remote sensing domain (comparison, logical verification and relational inference over small objects in aerial layouts), direct answering (S0) with the original ultra-resolution image under each model default preprocessing (V2); four-choice questions except anomaly detection, which asks for the grid coordinate of the one anomalous cell out of 100; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.

Top models

#ModelScore
1GPT-5.544.2
2GPT-5.435.8
3Gemini 3.1 Pro (Preview)34.5
4Qwen 3.5 27B32.5
5GLM-4.5V30
6Gemma 4 26B21.7

Interactive version: theaggregate.ai/benchmark?slug=ultravr-remote-sensing · How It Works · Data refreshed daily, snapshot 2026-09-29.