World Labs' Atlas builds large 3D worlds from a single image with camera-path video
theworldlabs · x · 2026-10-23
World Labs demonstrated Atlas, which lets you build large 3D environments starting from a single input image.
You can then define a camera path through the generated world and produce a video along it. The team showed how the 3D structure, camera path and final flythrough connect, outlining a full image-to-3D-world-to-video pipeline.
More from Multimodal
- AI Video for Less Than a Movie Ticket: Dreamina at $0.29 Plus CapCut PC Workflow — JaynitMakwana · 2026-10-23
- AI video hook technique: fill the frame with text before any footage appears — aftahi_ai · 2026-10-23
- Design explorations for Vercel's First of Kind season 3 revealed — soleio · 2026-10-23
- One person made 100+ product videos in a week with Claude, Figma and ElevenLabs — JosephJacks_ · 2026-10-23
- Seedance 2.5 demo: a world that lost its colors except one kept behind glass — LudovicCreator · 2026-10-23
- AI video: an orange cat in a parka rides a flying vacuum from the South Pole to the Amazon — legendofketty19 · 2026-10-23