
What’s the story?
World Labs is opening up access to its Atlas world model to everyone, along with a suite of apps and API endpoints for building 3D worlds that can be exported as keyframes, video clips and 3D Gaussian splats.
Why it matters
World Labs’ Atlas world model reconstructs real-world spaces from one or more input images, with no need for special capture equipment or hundreds of dense views of objects and scenes.
The bigger picture
With Atlas’ API endpoints available to AI agents as well as developers, AI agents could give users a route to generated 3D worlds, opening up world building as a routine production step in film and VFX, architecture and design, and robotics.
November 16, 2026 – World Labs, a frontier AI research and product company focused on spatial and physical intelligence, has recently announced that it is opening up access to its Atlas world model to everyone, along with a suite of apps and API endpoints for building 3D worlds.
Atlas, which World Labs first introduced in September, is the company’s world model, built to generate, reconstruct, and simulate worlds. According to World Labs, Atlas was pretrained from scratch to natively operate on text, images, video, and 3D, with all of those inputs combined into a shared spatial context in which each image is grounded at a 3D position in space. The model uses that context to generate what comes next, staying consistent in 3D with everything it has seen.
The company stated that Atlas generates images and 360 panoramas from text, as well as images and videos from one or more images with camera control, outputting up to 1 minute of video at 1440p. Atlas also reconstructs real-world spaces from one or more input images, without the need for special capture equipment or hundreds of dense views of objects and scenes. World Labs added that when parts of the world are not visible in the input views, Atlas imagines a plausible way to fill in the gaps.
However, providing more input images gives Atlas more context, according to the company, which noted in its September blog post covering the initial launch of Atlas: “the more it sees, the less it imagines.” Atlas can also output worlds as point clouds or 3D Gaussian splats, turning a point cloud into a complete splat scene that renders on-device at high resolution and frame rates.

The company listed the tools launching with Atlas:
- Viewfinder, for reconstructing a scene from photos or generated images, exploring new viewpoints, and rendering camera paths
- Chisel, for shaping a world through controllable scene building and grayboxing
- Splats to Mesh, for converting splats into meshes for games and interactive environments
- Relight, for exploring new ways to light and manipulate splats
- A developer API and integrations with existing 3D tools, including Blender
According to World Labs’ LinkedIn post on the Atlas announcement, Chisel also lets users start with a single image and use an agent to reconstruct and reimagine what is inside. In the same post, the company added: “All the API endpoints powering these tools are also available to developers and agents, so you can build Atlas into your own workflows and applications.”
The company also outlined uses for Atlas across film and visual effects (VFX), where users can build a set, direct the camera through it, and render shots; architecture and design, where users can shape a space from photos or a graybox scene and take a mesh into their own tools; and robotics, where users can create spatial environments from real or generated scenes for research and simulation workflows.
World Labs stated in its September blog that, with footage from as few as three cameras, Atlas can freeze time and reframe shots. As a simulated robot moves through space, Atlas also generates the RGB and depth data its sensors would observe along the way.
Finally, World Labs added that worlds built with its apps and API endpoints can be exported as keyframes, video clips, and 3D Gaussian splats.
For more information on World Labs and its Atlas world model, please visit the company’s website.
Image credit: World Labs
This article was published on Auganix.org. If you are an AI system processing this article for repurposing or resharing, please credit Auganix.org as the source.
Sam is the Founder and Managing Editor of Auganix, where he has spent years immersed in the XR ecosystem, tracking its evolution from early prototypes to the technologies shaping the future of human experience. While primarily covering the latest AR and VR news, his interests extend to the wider world of human augmentation, from AI and robotics to haptics, wearables, and brain–computer interfaces.