Skip to content

Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.

2.7 KB
Raw Blame

title: "readme-3d - interactive 3D models in your GitHub README" description: "Convert GLB, glTF, OBJ, and STL files into the ASCII STL markdown blocks GitHub renders as an interactive 3D viewer. CLI, library, tutorials, and AI-agent skills."

3D in markdown, the easy way

GitHub renders ASCII STL inside a markdown code fence as a live, draggable 3D viewer - in READMEs, issues, pull requests, discussions, and gists. No JavaScript, no images, no plugins.

readme-3d is the toolchain around that feature: it converts real models (GLB, glTF, OBJ, binary STL) into paste-ready blocks, simplifies meshes to fit GitHub's 512 KB markdown limit, and size-checks your files so they never silently stop rendering.

npx readme-3d model.glb >> README.md
npx readme-3d check README.md

See it live in the repo README → (GitHub renders the 3D viewer there; this docs site shows static markdown.)

Try it in your browser

Paste any ASCII STL into the playground to preview exactly what GitHub will render - or generate a model from a text prompt, free, at three.ws/forge.

Tutorials

  1. Hello, 3D markdown - your first embedded model in five minutes, no tools required
  2. Convert a GLB - the CLI end to end: convert, budget, tune
  3. Text → 3D → README - generate a model from a text prompt with the free three.ws Forge API and embed it
  4. Size budgets - the exact math of GitHub's 512 KB render limit
  5. Beyond the README - 3D in issues, PR reviews, discussions, gists, and printing projects

For AI agents

The repo ships a Claude Code skill and ready-made prompts so "add a 3D model of a fox to my README" is a one-line request. Model generation is available to any MCP-capable agent through the free three.ws forge_free tool.

Links