VLAN - Real World 30 Trajectories - ๐๐act๐๐ฏ๐ โ: leaderboard
Independent evaluation cohort. S4T7=30real-worldtrajectories; language/actionfamilymeans andharmonicAccSR. A4T14=fixed200testtrajectories, 2/scene, seed42, nineVLMs, identicalbudget; notinterchangeablewith1931fullsplit. FullsplitDroneSRact/GRinconsistency held separately. Metric: Native success score (0โ1). Source: https://arxiv.org/html/2611.18346v1
Metric: Native success score (0โ1). Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL3.5-8B | 0.617 | #1588 |
| 2 | Qwen 3 VL 8B Instruct | 0.483 | #1025 |
| 3 | Qwen 2.5 VL 7B Instruct | 0.467 | #1893 |
| 4 | Idefics3-8B-Llama3 | 0.417 | #5775 |
| 5 | Molmo2-8B | 0.35 | #1629 |
| 6 | Llama 3.2 11B Instruct | 0.217 | #5226 |
| 7 | Qwen 2.5 VL 3B Instruct | 0.083 | #2648 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=vlan-real-world-30-trajectories-act ยท How It Works ยท Data refreshed daily, snapshot 2026-11-17.