World Labs Unveils Atlas: A Native Multimodal World Model
gowthami_s · x · 2026-09-02
World Labs introduced Atlas, a next-gen world model and a multimodal autoregressive diffusion transformer natively operating on text, images, video, and 3D.
Core Capabilities:
- Camera-Controlled Gen: Generates up to 1-minute 1440p videos from images with pixel-perfect control.
- Spatial Reconstruction: Reconstructs 3D scenes from sparse inputs, outputting novel views and explicit 3D data.
- Space-Time Simulation: Models space-time for effects and robotics Real-to-Sim workflows.
- Image Gen: Creates images/360 panoramas from text, handling text rendering and various styles.
Performance scales with compute, powering future versions of Marble.
Related event: World Labs Launches Atlas, an Omni World Model for Spatial Intelligence(64 posts)→
More from Multimodal
- Apple's immersive horror short Remnants wows Vision Pro users with 3D presence — Scobleizer · 2026-10-18
- H3 360° Orbit Camera LoRA for MiniMax Open-Sourced, Generates Smooth 5.2s Shots — merica420_69 · 2026-10-18
- 'Tesla in Pakistan' AI video made with Seedance 2.5 goes viral — SimplyAnnisa · 2026-10-18
- Borges's nine stories turned into walkable 3D worlds with AI model Fable 5.1 — emollick · 2026-10-18
- Midjourney cozy couple photo: full prompt shared for killing the AI plastic look — tisch_eins · 2026-10-18
- Higgsfield playbook: scene templates, paid creator network, $20-to-enterprise funnel — sujingshen · 2026-10-18