Meshy entrusts 3D creation to a conversational agent

MeshyAI launches its Meshy 3D Agent beta, a conversational assistant that converts text or images into ready-to-use 3D models for Blender, Unity or Roblox.

MeshyAI is opening the beta for its Meshy 3D Agent, presented as the first AI agent dedicated to 3D creation. This approach breaks away from the logic of isolated prompts: the user interacts with the agent in their workspace, like a collaborator, starting from text or an image.

The agent begins with a framing phase, proposing several creative directions and asking questions about the intended use (3D printing, game assets, visual rendering). It then generates, in a single go, several 2D variations that can be refined by continuing the conversation. The chosen version is converted into a textured, ready-to-use 3D model, exportable in GLB or FBX formats to Blender, Unity, Unreal, or Roblox, or intended for printing. In parallel, it answers technical questions, from topology to optimization or print settings.

The official demonstration shows a photo transformed into a chibi figurine, available in several outfits then converted into a ready-to-print model, with the agent automatically managing supports and infill.

The agent is open to everyone during the beta, accessible from the Meshy workspace, but generations consume credits: 100 per month in the free version, 1,000 with the Pro offer at €19 per month, including a private license and advanced features. The team indicates that feedback from testers will guide future developments.