Shutterstock Unveils Revolutionary Generative 3D API

Shutterstock’s new generative 3D API, now available for enterprise use, produces realistic 3D models as well as 360 HDRi for rich and detailed natural environments for lighting 3D scenes.

By Brian K. Neal

Shutterstock, Inc., a global leader in creative content, has announced the launch of its groundbreaking generative 3D API. This innovative tool provides enterprises with a rapid and ethical solution for creating realistic 3D models using AI. The API is built on the multimodal NVIDIA Edify generative AI architecture and has been trained exclusively on Shutterstock’s vast library, including over half a million ethically sourced 3D models, more than 650 million images, and comprehensive Shutterstock metadata.

In addition to the generative 3D API, Shutterstock introduced an early access program for its 360 HDRi generator API. Also built on NVIDIA Edify, this API creates rich and detailed natural environments to enhance the lighting in 3D scenes. Both new offerings will be demonstrated live at the “Enhancing 3D Pipelines with Generative AI” session during the SIGGRAPH 2024 conference.

Dade Orgeron, Vice President of Innovation at Shutterstock, emphasized the transformative potential of these new capabilities. “Our generative 3D technology revolutionizes traditional processes, cutting down tasks that once took hours to mere minutes. This advancement not only accelerates 3D creation but also ensures fair compensation for original creators, who can choose to opt out of the training dataset,” Orgeron stated.

The expansion of 3D technology worldwide has heightened the need for ethical AI solutions. Shutterstock’s text-to-3D and image-to-3D generator API provides a timely response, enabling enterprises to accelerate production pipelines and populate virtual environments with ethically trained AI. The API supports seamless integration into any application or workflow, allowing users to quickly move from concept to creation. The process starts with a 2D preview image for user approval, followed by the generation of 3D files in GLB, USDz, and OBJ formats, complete with physically based rendering (PBR) materials.

Over the past 16 months, Shutterstock has refined this technology, resulting in significant improvements in quality and prompt adherence. The company is also committed to ethical compliance, legal integrity, and brand safety through tailored protections and content controls. Shutterstock has implemented robust safeguards to prevent the generation of inappropriate content and protect against the misuse of major brand intellectual property.

To assist users in adopting this technology, Shutterstock offers intuitive documentation, making it accessible even to non-developers. Lightweight plugins and code samples for tools like Blender and Gradio further streamline integration.

At SIGGRAPH 2024, early beta testers such as HP, Trigger XR, and WPP will showcase their use cases for the generative 3D API. HP will demonstrate 3D printing and physical prototyping, while Trigger XR will highlight asset creation for XR environments using an Apple Vision Pro. WPP will explore generative 3D for virtual production on the NVIDIA Omniverse platform. Shutterstock’s in-house production service, Studios, will also present their immersive production capabilities and how they leverage this technology for commercial clients.

The generative 3D API is now available to enterprise customers, with a self-serve option coming soon. Generative credit packs start at $25, making this cutting-edge technology accessible to companies of all sizes.