Multimodal world creation
Create spatially consistent 3D worlds from text, images, videos, or 360 panoramas, giving users multiple starting points for generation.
World Labs is a spatial intelligence company. Its first product, Marble, creates persistent 3D worlds from text, images, videos, or 360 panoramas.
World Labs is a spatial intelligence company focused on frontier world models and its first product, Marble. Marble generates spatially consistent, high-fidelity, persistent 3D worlds that users can move through, edit, and inhabit. The product is positioned around turning prompts and media inputs into usable 3D spaces rather than static images.
The site frames World Labs as a frontier AI research and product company working on models that can perceive, generate, reason, and interact with the 3D world. Its stated aim is to support use cases across storytelling, design, simulation, and immersive experiences in both virtual and physical settings.
Create spatially consistent 3D worlds from text, images, videos, or 360 panoramas, giving users multiple starting points for generation.
Use Chisel-style 3D layout control to precisely shape the arrangement of objects and space inside generated worlds.
Edit specific elements or reshape the entire environment while keeping the world persistent and navigable.
Expand, edit, and combine generated worlds to build larger environments instead of treating each scene as a one-off output.
Download and export in multiple 2D and 3D formats for downstream workflows and pipelines.
Browse Marble Labs for case studies, tutorials, documentation, and community examples that show how the product is used in practice.
Creators can turn prompts or reference media into explorable 3D environments for narrative scenes, concept worlds, and visual development.
Designers can generate a space, then adjust layout and individual elements to study compositions or presentation-ready scenes.
Teams working on simulation or embodied AI can use persistent worlds as editable environments for testing and training workflows.
Product and creative teams can expand or combine generated worlds to build larger environments for immersive experiences or virtual production.
Developers and hobbyists can use Marble Labs tutorials, documentation, and community projects to learn the workflow and adapt examples to their own builds.
Marble is World Labs’ first product and is described as generating spatially consistent, high-fidelity, persistent 3D worlds from a single image, video, text prompt, or other multimodal inputs.
The site presents Marble Labs as a place to explore case studies, tutorials, documentation, and community projects built with Marble and World Labs technology.
The source highlights downloadable and exportable outputs in various 2D and 3D formats, but it does not list exact file types or a complete export matrix.
A separate pricing page exists on the domain, but the fetched /pricing page currently returns a 404, so no verified plan or price details are available from the provided sources.
The provided sources mention tutorials, documentation, a Discord community, and a Gaussian Splatting Three.js plugin under Marble Documentation and Learn, but they do not describe a full public API or integration catalog.
Endless Tools is a no-code web design tool for 3D assets, motion content, and visual effects, built for designers, marketers, and creators.
Avaturn creates realistic 3D avatars from a selfie, with customization, export, and embedding for apps, games, metaverse experiences, and campaigns.
Rendair AI is a web-based AI workspace for architects and designers to create, edit, upscale, and animate architectural visuals from text, sketches, plans, CAD, and 3D models.
Kaedim is an AI-powered 3D asset production service for studios and brands, turning sketches, references, product photos and briefs into production-ready 3D assets.
Realibox is an AI+3D creative visual design platform for 3D design, AI rendering, CMF color generation, packaging visualization, and commercial model asset downloads.
Tripo 3D is an AI 3D modeling platform that turns text, images, sketches, and multi-view inputs into 3D assets. It includes a browser-based studio, an API for automation, and export support for GLB, STL, and OBJ.