RunwayML updates Gen-Render with 3D-aware text-to-image models and asset export

AI ยท 4 min read

RunwayML updates Gen-Render with 3D-aware text-to-image models and asset export

RunwayML upgraded Gen-Render to produce 3D-aware renders from text prompts that include implicit depth, normal maps, and albedo layers. Designers can ask for "a chrome kettle on a wooden table from a 45-degree angle" and receive not only a composited image but also depth and material passes useful for compositing and AR previews.

The system can export assets as layered PSDs, PBR texture maps, or OBJ/GLTF + textures for rapid prototyping in AR or 3D scenes. RunwayML added a material editor so designers can tweak reflectivity and roughness before export.

Runway positions this update for product designers and advertising teams who need realistic assets fast, noting that the model can be fine-tuned on brand materials to match product finishes accurately. The feature is available to subscribers with high-tier compute plans.