Nimbus AI unveils Pianta: a 3D-aware text-to-scene generator with live collaboration

AI · 6 min read

Pianta converts natural language prompts into an editable 3D scene graph and runs in the browser with GPU-backed rendering. Users can adjust materials, scale objects, and re-sample lighting while the tool maintains semantic consistency. The company highlights tight integration with Blender and Unity through one-click export.

Alongside the launch, Nimbus AI revealed a $35 million Series A led by Frontier Ventures, bringing total funding to $48 million. Investors pointed to Pianta's multimodal backbone and developer APIs as differentiators that can accelerate virtual production and rapid prototyping workflows.

Early customers include two mid-size studios and an architecture firm piloting Pianta for rapid site studies. Nimbus plans to roll out a plug-in SDK for on-premise deployment, optimized path tracing, and enterprise collaboration controls over the next nine months.