Button with the text '3D MODELS' in white on a black background with rounded edges.
  • Display of various toy figurines and decorative objects on circular platforms.

    Stable Point Aware 3D

    Stable Point Aware 3D (SPAR3D) introduces real-time editing and complete structure generation of a 3D object from a single image in less than a second.

    With a first-of-its-kind architecture, SPAR3D combines precise point cloud sampling with advanced mesh generation to deliver unprecedented control over 3D asset creation. 

    Read the Technical Report

  • A Stability AI branded AI training device with a raccoon cartoon mascot and a small platform for AI training.

    Stable Fast 3D

    Stable Fast 3D generates high-quality 3D assets from a single image in just 0.5 seconds, with significant architectural improvements and enhanced capabilities based on TripoSR.

    It serves game and VR developers, and professionals in retail, architecture, design, and other graphic-intensive fields.

    Read the Technical Report

  • Animation showing a 3D model of a bear with labeled 4D outputs and a smaller inset of the bear, along with a film camera icon.

    Stable Video 4D

    Stable Video 4D is an innovative model that allows users to upload a single video and receive dynamic novel-view videos from eight different angles. It transforms a single object video into multiple novel-view videos, generating 5 frames across 8 views in about 40 seconds with a single inference.

    Users can specify camera angles, tailoring the output to meet specific creative needs, delivering a new level of versatility and creativity.

    Read the Technical Report

  • Collection of small animal and character figurines on circular bases, including dogs, cats, a pig, and various characters, arranged in rows.

    Stable Video 3D

    Description goLeveraging the power of Stable Video Diffusion, SV3D sets a new benchmark in 3D technology by ensuring superior quality and consistency in novel view synthesis. 

    This model comes in two distinct variants: SV3D_u, producing orbital videos from a single image, and SV3D_p, which offers enhanced capabilities for creating full 3D videos from both single and orbital imagery.

    Read the Paper

  • Comparison of Zero123-XL and Stable Zero123: each row shows a plush bear, a sandwich, and a sofa, with the Zero123-XL items in the top row and the Stable Zero123 items in the bottom row.

    Stable Zero 123

    Stable Zero123 is an advanced AI model specialized in generating 3D objects. It stands out due to its capability to accurately interpret how objects should appear from various perspectives, which is a significant advancement in the realm of 3D visualization.

  • Comparison of two models, OpenLRM and TripoSR, with images of a meditating person, a burger, and a horse.

    Stable TripoSR

    In collaboration with Tripo AI, we're excited to introduce TripoSR, our rapid 3D object reconstruction model. This new image-to-3D model is designed to cater to the growing demands of entertainment, gaming, industrial design, and architecture professionals, with responsive outputs for visualizing detailed 3D objects.

Build with Stability AI 

Integrate our state-of-the-art models into your infrastructure with Self-Hosted License or use our Stability AI API to power your applications.