Text to 3D AI – Open AI Launches Shape-E
3D modeling is constantly evolving with the introduction of AI tools from startups as well as tech giants, and OpenAI’s Shap-E adds to a number of recent announcements in this space. This tool generates 3D models from text and images, suggesting how 3D content will be created in the not-so-far future.
OpenAI has recently launched Shap-E, a tool that can generate 3D models from images or text. With a neural network trained on millions of 3D shapes and natural language descriptions, Shap-E is able to create realistic and diverse 3D models from a simple text prompt. This free tool is available on GitHub and runs smoothly on the user’s own PC. Shap-E is both faster and more accurate than OpenAI’s own Point E, another text-to-3D AI model. As part of a multi-billion dollar partnership between Microsoft and OpenAI, Shap-E is powered by Microsoft’s Azure cloud infrastructure and is the latest innovation from OpenAI, which is also responsible for ChatGPT and Dall-E. Whether for education, entertainment or design, Shap-E is a powerful and exciting tool for creating 3D models, without manual sculpting.
In this space, Magic Fabric has previously covered other exciting services like Kaedim and Luma AI. Kaedim produces game assets using image input with the help of neural networks. Luma AI generates 3D models from photos or videos using a similar approach but with image input from mobile phones.