LinkS2Bench: leaderboard
LinkS2Bench (LinkS²Bench, 2026) tests whether vision-language models can anchor dynamic drone-video observations to static satellite maps, over 17,903 questions in 12 perception, localization, relation and reasoning tasks. Score is the mean of the 12 task accuracies (%); higher is better.
Metric: Average (%) of the 12 task scores (accuracy, MRA for zone counting, ACC@1s for temporal grounding) 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 October 2027. 18 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 51.1 | #102 |
| 2 | Seed 1.8 | 47.8 | #258 |
| 3 | GPT-5.4 | 46.6 | #138 |
| 4 | Qwen 3.5 397B A17B | 45.6 | #270 |
| 5 | GLM-4.6V-Flash | 44.3 | #1049 |
| 6 | Qwen 3.5 35B A3B | 44.1 | #494 |
| 7 | Gemini-3.1-Flash [LinkS2Bench checkpoint unspecified] | 42.8 | |
| 8 | Kimi K2.5 | 38.4 | #272 |
| 9 | GPT-5.4 Mini | 38.2 | #409 |
| 10 | Claude Sonnet 4.6 | 37.7 | #179 |
| 11 | LLaVA-OneVision-1.5-8B | 33.7 | #1876 |
| 12 | LLaVA-OneVision-1.5-4B | 33.2 | #1874 |
| 13 | Qwen 3.5 9B | 30.7 | #838 |
| 14 | Llama 4 Scout | 26.6 | #1825 |
| 15 | Qwen 3.5 4B | 23.7 | #1270 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=links2bench · How It Works · Data refreshed daily, snapshot 2026-11-22.