Global

text-to-cad is a library of agent skills for generating, inspecting, sourcing, slicing, and handing off CAD and robot-description artifacts from local project files.

Disclaimer: This is a third-party integration and not made by Blender

AssetFurnace: the most complete open-source image-to-3D pipeline (Oct 2026). Prompt or picture → textured, rigged, animated game character, locally on Apple Silicon or NVIDIA. 2, Hunyuan3D, SF3D, auto-rig and 21 moves.

T RELLIS is a large 3D asset generation model. It takes in text or image prompts and generates high-quality 3D assets in various formats, such as Radiance Fields, 3D Gaussians, and meshes.

> Join our Wechat and Discord group to discuss and find help from us.

The five released Kimodo motion checkpoints accept either a UTF-8 prompt or a precomputed LLM2Vec embedding and generate local rotations plus root translations on CPU or Vulkan:

The output is not a point cloud or a soup of triangles. It is source code you can open, edit, and re-compile — optionally with per-part PBR materials (--paint) and articulation exported to OpenUSD/URDF (--motion).

Amagine3D is the open-source 3D capability layer Amagine is developing for hardware creation.

You can also adjust the sweep configuration file if you need to perform a hyperparameter search.