Thrixel, founded by University of Chicago computer scientist Rana Hanocka, launched a 3D creation platform that generates assets built from individually named parts with logical hierarchy, so AI agents can identify, assemble, and animate them directly in industry-standard game engines.

“AI has transformed text, images, video, and code, but 3D has historically remained locked behind slow, expensive, and highly specialized tools,” said Rana Hanocka, founder and CEO of Thrixel, in a press release.

Building production-ready 3D content still demands deep technical expertise, which puts it out of reach for many developers and slows down even large studios. Thrixel’s part-aware generative architecture produces structured 3D assets with individually named, identifiable components built for AI-driven pipelines. A car generated with Thrixel arrives with its wheels, doors, and chassis as separate named parts that can be animated or swapped.

Creators can manipulate individual parts directly with code, assemble complex assets from components, and bring them into existing 3D workflows. The platform combines vision language models, agentic pipelines, generative diffusion, and optimization techniques, each mapped to a discrete step in the 3D modeling pipeline, so creators can use the full workflow or apply specific steps to refine a model.

Thrixel grew out of years of research at the intersection of AI and computer graphics led by Founder and CEO Rana Hanocka, Assistant Professor of Computer Science at the University of Chicago. Her research laid key foundations for 3D AI, anchored by MeshCNN, a pioneering framework for deep learning on meshes, and Text2Mesh, an early milestone in text-driven 3D manipulation.

Source link