ProductsNEWS BRIEFING

World Labs Opens Atlas to Everyone with Viewfinder, Chisel, and Developer API

2 min read

88Z AI, 13 November — Fei-Fei Li’s World Labs has publicly opened access to its next-generation world model, Atlas. Users can now generate 3D Gaussian splats from images via the Viewfinder web app or create explicit 3D geometry using agentic workflows in Chisel. The company also released all underlying API endpoints, enabling developers to build custom tools or allow agents to interact directly with the models, with outputs including keyframes, video clips, and 3D Gaussian splats.

At a glance

  • World Labs opens public access to Atlas, a world model for consistent 3D environment generation
  • Viewfinder app reconstructs 3D Gaussian splats from single or up to 64 input images
  • Chisel app uses agentic workflows to create explicit 3D geometry from text or image references
  • Full API endpoints released for developers to build custom tools and integrate agents
World Labs Atlas platform interface showing Viewfinder and Chisel app entries
World Labs Atlas platform interface showing Viewfinder and Chisel app entries

Full Public Access and Core Capabilities

On November 12, 2026, World Labs announced it is opening access to its next-generation world model, Atlas, to everyone. Atlas is designed to generate and reconstruct consistent 3D environments. This release includes not just the model but also applications and API endpoints built to help users utilize Atlas in diverse ways. Users can transform ideas for building 3D worlds into keyframes, video clips, or 3D Gaussian splats. The company provides a dedicated landing page for getting started, offering immediate hands-on experience via web apps or access to developer documentation and model endpoints for technical integration.

Viewfinder: Spatial Intelligence for Scene Reconstruction

Viewfinder is a web application that leverages Atlas’s spatial intelligence to create worlds from images. It supports two primary modes: generating an entire world from a single image, or reconstructing real-world environments using up to 64 input images. This multi-view input capability enhances the accuracy and detail of reconstructions. Upon completion, users can export spatially consistent images, videos, or 3D Gaussian splats. This feature lowers the barrier to extracting 3D spatial information from 2D visual data, catering to workflows that require rapid scene digitization.

Chisel: Agentic Workflows for Explicit Geometry

Unlike Viewfinder, Chisel focuses on creating explicit 3D geometry and is built around agentic workflows. Users can input text prompts or reference images to construct their world in 3D space. Chisel allows exploration of various visual styles during creation, with final outputs rendered as images, videos, or exported splats. The agentic workflow design enables the system to intelligently handle complex 3D modeling tasks, translating natural language or visual references into structured 3D assets. This provides content creators with greater flexibility and control over the generation process.

Developer API and Ecosystem Integration

In addition to end-user web apps, World Labs has released all API endpoints powering these applications. This move targets developers and agents, enabling them to build custom tools that incorporate Atlas. Developers can use their agents to interact directly with the models, embedding 3D generation capabilities into broader workflows or automated systems. By opening its API, World Labs aims to foster third-party innovation based on Atlas, allowing more professional tools to leverage its consistent 3D environment generation technology and expanding the model’s boundaries in industry applications.

Sources & editorial note

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.

  1. @theworldlabsSource published

Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.

All AI news

Related reading

All stories
88Z.ai