AirGroundBench - Bearing Perception: leaderboard
Metric: Accuracy (%) on the bearing-perception VQA task; higher is better. Source: arxiv.org. Saturation forecast: Around 2034. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 Plus | 48.28 |
| 2 | Qwen 3.5 397B A17B | 45.26 |
| 3 | Qwen 3.5 27B | 44.83 |
| 4 | Qwen 3 VL 235B A22B | 41.38 |
| 5 | Kimi K2.5 | 38.63 |
| 6 | Seed 2.0 Pro | 37.93 |
| 7 | ERNIE 5.0 | 37.93 |
| 8 | GLM-4.6V | 27.59 |
| 9 | Gemini 3.1 Pro (Preview) | 20.69 |
| 10 | GPT-5.2 | 20.69 |
| 11 | Claude Sonnet 4.6 | 17.24 |
Interactive version: theaggregate.ai/benchmark?slug=airgroundbench-bearing-perception · How It Works · Data refreshed daily, snapshot 2026-09-29.