DriveSpatial - Understanding RMSE: leaderboard
DriveSpatial (2026) tests whether vision-language models can build and reason over a coherent driving scene from synchronized multi-view camera frames, with human-verified questions generated from scene graphs of nuScenes, Waymo, TruckScenes, AV2 and ONCE. This board covers the numeric spatial-understanding tasks on the full 15,670-question set, scored as root mean squared error of numeric answers such as distances and counts; lower is better.
Metric: Root mean squared error of numeric answers (distances, counts and other spatial measurements) on the spatial-understanding tasks on the full benchmark (15,670 human-verified questions over 20 tasks from nuScenes, Waymo, TruckScenes, AV2 and ONCE); open-source generalist and driving or spatial specialist VLMs; lower is better. Source: arxiv.org. Saturation forecast: Around May 2027. 12 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL3.5-8B | 14.26 | #1618 |
| 2 | Qwen 3 VL 8B | 14.32 | #927 |
| 3 | Gemma 3 12B (IT) | 14.65 | #1907 |
| 4 | DeepSeek-VL2-Small | 15.15 | #2523 |
| 5 | DriveSpatial SenseNova-SI (checkpoint unspecified) | 15.25 | |
| 6 | RoboTron-Drive | 15.49 | |
| 7 | LLaVA-Video-7B | 15.68 | #2303 |
| 8 | LLaVA-OneVision-7B | 15.76 | #3002 |
| 9 | InternVL3-8B | 15.95 | #1795 |
| 10 | DriveSpatial Ego3D (checkpoint unspecified) | 15.95 | |
| 11 | SpaceThinker | 16.96 | #1467 |
| 12 | LongVA-7B | 20.26 | #3480 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=drivespatial-understanding-rmse · How It Works · Data refreshed daily, snapshot 2026-11-22.