Introducing Stable Fast 3D: Rapid 3D Asset Generation From Single Images

Key Takeaways:

  • Stable Fast 3D generates high-quality 3D assets from a single image in just 0.5 seconds.

  • Built on the foundation of TripoSR, Stable Fast 3D features significant architectural improvements and enhanced capabilities.

  • The model has applications for game and virtual reality developers, as well as professionals in retail, architecture, design and other graphic-intense professions.

  • The model is available on Hugging Face and is released under Stability AI Community License.

  • Access the model easily on Stability AI API and on Stable Assistant chatbot and share your 3D creations in a 3D viewer and play with them in Augmented Reality. Get started with a free trial. 

We are excited to introduce Stable Fast 3D, Stability AI’s latest breakthrough in 3D asset generation technology. This innovative model transforms a single input image into a detailed 3D asset in just 0.5 seconds, setting a new standard for speed and quality in the field of 3D reconstruction.

How It Works

Users start by uploading a single image of an object. Stable Fast 3D then rapidly generates a complete 3D asset, including:

  • UV unwrapped mesh

  • Material parameters

  • Albedo colors with reduced illumination bake-in

  • Optional quad or triangle remeshing (adding only 100-200ms to processing time)

Here is a video describing how the model works and the improvements vs previous models.

Stable Fast 3D's unprecedented speed and quality make it an invaluable tool for rapid prototyping in 3D work, catering to both enterprises and indie developers in gaming and virtual reality, as well retail, architecture and design.

You can also use the model easily on Stability AI API and on Stable Assistant chatbot where you can share your 3D creations in a 3D viewer and play with them in AR (on Augmented Reality (WebXR) compatible devices).

Use Cases

Stable Fast 3D has several use cases in gaming and movie production. 

  • Use the fast inference time during pre-production, where experimentation is key

  • Static assets for games (Background objects, clutter, furniture)

  • 3D models for E-commerce

  • Fast model creation for AR/VR

Speed Meets Quality

Stable Fast 3D outperforms competitors in several key areas:

  • Unmatched speed: 0.5 seconds per 3D asset generation on a GPU with 7GB VRAM or in close to a second on Stability AI API

  • High-quality UV unwrapped mesh and material parameters

  • Reduced illumination entanglement in textures

  • Ability to generate additional material parameters and normal maps

Compared to our previous SV3D model, Stable Fast 3D offers dramatically reduced inference times - 0.5 seconds versus 10 minutes - while maintaining high-quality output.

Research and Development

Stable Fast 3D builds upon our previous work with TripoSR, but features a completely retrained model with significant architectural changes. These improvements allow for explicit mesh generation and incorporate novel techniques for fast textured mesh creation.

A technical report on Stable Fast 3D is available here and highlights how we achieve fast inference speeds with reduced baked illumination and material parameters.

Availability

Stable Fast 3D model code is available on Github and model weights and demo space on Hugging Face. Stable Fast 3D is released under Stability AI Community License that allows non-commercial use and commercial use for individuals or organizations with up to $1M in annual revenue. Above this threshold, contact us for Enterprise Licenses.

The model is also accessible via our API, and on Stable Assistant.

The Project page is available here.

Image Licenses

Images in this blog post taken from GSO (License), OmniObject3D (License), as well as Stable Assistant.

To stay updated on Stable Fast 3D and other Stability AI developments, sign up to our Newsletter, follow us on Twitter, Instagram, LinkedIn, and join our Discord Community.

Previous
Previous

Stability AI names Hanno Basse as new Chief Technology Officer

Next
Next

Introducing Stable Video 4D, Our Latest AI Model for Dynamic Multi-Angle Video Generation