World Labs unveils Atlas, an omni world model for spatial intelligence with 1440p generation
theworldlabs · x · 2026-10-16
World Labs has introduced Atlas, its next-generation omni world model, with beta access now rolling out.
- Architecture: a multimodal autoregressive diffusion transformer pretrained from scratch to natively handle text, images, video, and 3D, combining all inputs into a shared spatial context for 3D-consistent generation.
- Capabilities: pixel-perfect camera-controlled image/video generation (up to 1 minute at 1440p); spatial reconstruction from 1 to dozens of images producing explicit 3D outputs, beating specialized 3D reconstruction SOTA; space-time simulation enabling reframing and Real-to-Sim robotics workflows; text-to-image and 360 panoramas.
- Scaling: performance improves with training compute, a trend the team expects to hold.
Atlas will power future versions of Marble and other products; the site shows pixel-precise camera path control demos.
Related event: World Labs Unveils Atlas World Model with Pixel-Level Camera Control(2 posts)→
More from Multimodal
- One JSON UI prompt, three contenders: Claude vs ChatGPT vs Figma Make for iOS mockups — Tegadesigns · 2026-10-16
- Opus 5.5 + Midjourney: 'models have taste, you just point them right' — yuwen_lu_ · 2026-10-16
- Gemini 4 Argon hits AssemblyAI's LLM Gateway, tops voice bench at 99.5 across 50+ models — AssemblyAI · 2026-10-16
- Envato Launches AI Creative Toolkit Bundling Nano Banana, Veo and More — alifcoder · 2026-10-16
- Musk amplifies Grok Imagine dance video demo tackling generative AI's hardest motion — elonmusk · 2026-10-16
- CameraVLM: 9B model cracks camera-motion understanding where most VLMs flip coins — tencent · 2026-10-16