LinkS2Bench - Trajectory Matching: leaderboard
Metric: Trajectory Matching: multiple-choice accuracy (%) on which satellite trajectory a moving target in the video follows on the full LinkS2Bench (17,903 questions over 1,500 UAV videos paired with satellite images), temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around August 2027. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 397B A17B | 50 |
| 2 | Seed 1.8 | 47.3 |
| 3 | Gemini 3.1 Pro (Preview) | 44.7 |
| 4 | GPT-5.4 | 43 |
| 5 | Qwen 3.5 35B A3B | 42.6 |
| 6 | GPT-5.4 Mini | 33 |
| 7 | Claude Sonnet 4.6 | 31.3 |
| 8 | Kimi K2.5 | 26.1 |
| 9 | Llama 4 Scout | 22.6 |
| 10 | Ministral 3 8B | 19.3 |
| 11 | Ministral 3 14B | 18.2 |
| 12 | Qwen 3.5 9B | 15.3 |
| 13 | Ministral 3 3B | 14.5 |
| 14 | Qwen 3.5 4B | 7.3 |
Interactive version: theaggregate.ai/benchmark?slug=links2bench-trajectory-matching · How It Works · Data refreshed daily, snapshot 2026-10-07.