Press Kit

Yugma is an AI-native 3D studio where you describe scenes in natural language and watch them build in real time.

Yugma is a browser-based 3D creation studio that puts AI at the center of the workflow. Instead of manually placing and configuring objects, users describe what they want in plain language, and Yugma's AI Director composes the scene using 15 structured tools in an agentic loop. The result is production-ready 3D that can be exported as GLB, shared via live links, or embedded anywhere — all without downloads, plugins, or prior 3D experience.

Yugma is an AI-native 3D studio designed for a world where natural language is the most powerful creative tool. Built entirely in the browser with React Three Fiber and powered by Anthropic's Claude, Yugma lets users describe 3D scenes in plain English and watch them assemble in real time. The system goes far beyond simple text-to-3D generation: it uses an agentic AI loop with 15 structured tools that iteratively creates, positions, materials, lights, and composes objects into coherent, spatially aware scenes.

At the core of Yugma is the AI Director — a scene composition engine that understands spatial relationships, material physics, and design principles. When a user says "Build a sci-fi hangar with neon lighting," the Director doesn't just place random objects. It creates walls with correct dimensions, positions lights at architectural anchor points, applies emissive materials for the neon glow, and groups related elements — all through a multi-iteration tool-use loop where each step builds on the last.

Yugma also supports real-time collaboration with live cursors and room codes, text-to-3D mesh generation via Meshy.ai, 30 PBR material presets, an AI material inspector, GLB export, screenshot capture, and embeddable 3D viewers. The platform is designed for game developers, architects, e-commerce teams, educators, and anyone who needs 3D without the steep learning curve of traditional tools. It is currently in public beta at yugma.studio.