3DFY lets devs create 3D models based on text prompts

3DFY.ai announced the launch of 3DFY Prompt, a generative AI that lets developers and creators build 3D models based on text prompts.

The Tel Aviv, Israel-based company said the tech democratizes professional-quality 3D model creation, enabling anyone to use text prompts to create high-quality models that can be used in gaming, design or virtual environments.

Until now, high-quality 3D model creation required professional labor and lots of it. Some recent advances in AI, such as those recently announced by Nvidia, Google and OpenAI, offer effortless 3D model creation. But 3DFY said those solutions severely compromise asset quality, which, in practice,
hinders the assets’ usability.

To be more than a cool fad, generative AI models need to be tangibly useful, 3DFY said.

3DFY’s text-to-3D technology uniquely offers a scalable AI-based solution which generates 3D models comparable to those produced by human 3D modelers. The tradeoff for the professional model quality comes in the form of a category-bound solution, meaning only objects within a range of defined classes can be generated.

However, with a continuously expanding range of available categories, this limitation will quickly diminish with time. The result is beautiful 3D models, which are always divided into semantically meaningful parts,
constructed with professional-level mesh topology available in multiple levels of detail with addition of physically based rendering textures at any desired resolution.

In addition, 3DFY.ai employs strictly ethical AI, as the vast amounts of training data utilized to train the company’s AI are 100% synthetic and generated in-house using advanced computer graphics methods.

Eliran Dahan, CEO of 3DFY, said in a statement, “With 3DFY Prompt, 3D creation can now be available to everyone, regardless of budget or experience, in ways that enable self-expression and save time, money and manual work for creators and without having to worry about possible copyright issues. We’re already working on additional object categories and new, unique functionality. Our overarching vision is to make 3D authoring as fun and easy as playing an online game. We want everyone to be empowered to create 3D content.”

3DFY Prompt is live, and can be experienced for free right now. You can play around, creating your own previously unexpressed desires (“an ottoman that looks like an Oreo” anyone?) and save them for later.

The company has raised $3 million in a pre-seed round to date, and it has seven people, with experience in computer vision, graphics, and AI and machine learning. Dahan started the company in 2019 with cofounder Tal Kenig.

In an email to GamesBeat, Dahan said, “[We] go way back and worked together for many years developing medical imaging technologies. As radiology veterans, we are 3D natives, so when we learned how much manual labor is involved in the production of computer graphics 3D assets, we immediately knew that’s where we want to make an impact. Ever since we started out, we are pursuing a clear vision: To make 3D model creation simple and easy for everyone, without compromising on quality. Somewhat similar to what Canva has done for graphic design. What you see now as the first version of 3DFY Prompt is just the tip of the iceberg, as we are working on additional, unique functionality, materializing this vision one step at a time.”

Asked if the AI could eliminate jobs of game developers, Dahan said, “We think we will see a trend here, similar to other generative AI fields, where creators would require less technical prowess and more skills around creatively using the new capabilities. But in any matter, it will be humans working with machines.”