Stability AI, a leading generative AI startup, has shared a sneak peek at major upgrades coming to their AI image generation offerings, headlined by Stable 3D, which automates creating 3D models from text or images.
Now available in private preview, Stable 3D eliminates much of the complexity of 3D modeling, enabling users to generate draft quality 3D objects in minutes versus hours or days traditionally. By inputting text prompts or images, even non-experts can rapidly produce models. Stability AI claims this breakthrough democratizes 3D content creation by vastly increasing speed and reducing costs.
The 3D files that are generated can be exported in standard .obj format and refined in 3D software like Blender and Maya, or imported in a game engine, such as Unreal Engine 5 or Unity. Stability AI positions Stable 3D as empowering independent artists, designers, and developers to generate thousands of 3D objects daily.
Other previews included Sky Replacer for quickly swapping skies in images and Stable FineTuning which will allow enterprises and developers to quickly fine-tune pictures, objects, and styles. The company also announced the integration of Content Credentials and invisible watermarking for images generated via the Stability AI API.
While light on technical details, the previews highlight Stability AI's intent to continue pushing the boundaries of AI content generation for different media types. In September the company launched Stable Audio, which was recently selected as one of TIME's Best Inventions of 2023. The tool enables users to use AI to generate original music and sound effects through a user-friendly interface, responding to prompts to create diverse musical compositions.
While still early days, the potential implications of technologies like Stable 3D on media creation could be far-reaching. The company's expansive family of generative AI models, which now span audio, image, language and 3D, signal a commitment to pushing boundaries. As Stability AI continues enhancing capabilities across modalities, they could make advanced creative tools accessible to professionals and enthusiasts alike.
This ethos of democratization is central to their mission. Offerings like Stable Diffusion and Stable Audio already unlocked AI art and music generation for many users. Stable 3D points to similar disruption in 3D content creation. Looking ahead, integration between these technologies could hint that significantly more capable multimodal AI models could be on the horizon for the ambitious startup.