Atlas
Atlas is a multimodal autoregressive diffusion transformer for generating, reconstructing and simulating space. It generates 1 minute of 1440p camera-controlled video.
One model for camera-controlled video, sparse-view 3D reconstruction (from one to dozens of images), space-time simulation and 360 panoramas; World Labs reports 75-94% human preference over specialized video models on camera following and scaling with compute. Early access with partners.
- Date
- Tuesday, 1 September 2026
- Lab
- World Labs
- Kind
- model
- Access
- research preview
Figures
| Measure | Value | Measured by |
|---|---|---|
| Human preference on camera-following | 75-94% win rate vs five specialized video models (75% to 94%); World Labs' own evaluation | company |
Sources
This record was checked and corrected against its sources on 6 October 2026. How we check