Repolode
DayuanJiang

DayuanJiang/next-ai-draw-io

A web, desktop and MCP tool that draws and edits draw.io diagrams from natural-language chat with an AI model.

Open-source alternative to:Lucidchart

Stars
36,331
Forks
3,848
Open issues
38
Last commit
yesterday

What is it?

Next AI Draw.io lets you describe a diagram in plain language and have an AI model draw it directly on a draw.io canvas. You can also edit by hand as with any draw.io file, or select shapes and ask the AI to change only those. Each AI change is saved as a version that can be compared, restored or undone, and diagrams export as .drawio, .png or .svg.

The project ships as a web app, a desktop app for Windows, macOS and Linux, and an MCP server that lets agents such as Claude Code, Cursor and VS Code draw diagrams. It supports 25 model providers, including Bedrock, OpenAI, Anthropic, Ollama and any OpenAI-compatible endpoint, with API keys kept in the browser. It can be self-hosted locally, on Vercel, Cloudflare Workers or Docker, or deployed offline.

Key features

  • 01Generates architecture diagrams, flowcharts and sequence diagrams from a sentence, with AWS, Azure, GCP and Kubernetes icon libraries.
  • 02Edits existing diagrams by chat, highlighting the shapes the AI changed.
  • 03Lets the AI modify only the shapes you select on the canvas.
  • 04Keeps a version history of AI changes with thumbnails, comparison, restore and undo.
  • 05Exports to .drawio, .png, .svg and .drawio.svg.
  • 06Works as a web app, desktop app or MCP server for Claude Code, Cursor and VS Code.

Use cases

  • Sketching cloud architecture diagrams from a written description during design reviews.
  • Turning a screenshot or sketch into an editable draw.io diagram.
  • Letting an AI coding agent produce and check diagrams while working in the editor.
  • Drawing diagrams with a local model through Ollama for offline or intranet use.

Quick start

  1. 1

    Clone the repository

    git clone https://github.com/DayuanJiang/next-ai-draw-io
  2. 2

    Enter the project folder

    cd next-ai-draw-io
  3. 3

    Install dependencies

    npm install
  4. 4

    Copy the env file and add your provider and API key, then start the dev server and open localhost:6002

    cp env.example .env.local

Things to know

  • Small models tend to produce broken diagrams, so a capable model is needed.
  • Running it yourself requires configuring a model provider and API key.

Topics

FAQ

What is next-ai-draw-io?
A web, desktop and MCP tool that draws and edits draw.io diagrams from natural-language chat with an AI model. Next AI Draw.io lets you describe a diagram in plain language and have an AI model draw it directly on a draw.io canvas. You can also edit by hand as with any draw.io file, or select shapes and ask the AI to change only those. Each AI change is saved as a version that can be compared, restored or undone, and diagrams export as .drawio, .png or .svg.
Is next-ai-draw-io free to use? What license does it have?
next-ai-draw-io is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is next-ai-draw-io still maintained?
Actively maintained. The last commit to next-ai-draw-io was on Oct 19, 2026.
What can I use next-ai-draw-io for?
Sketching cloud architecture diagrams from a written description during design reviews. Turning a screenshot or sketch into an editable draw.io diagram. Letting an AI coding agent produce and check diagrams while working in the editor. Drawing diagrams with a local model through Ollama for offline or intranet use.
How do I install next-ai-draw-io?
Clone the repository: git clone https://github.com/DayuanJiang/next-ai-draw-io Enter the project folder: cd next-ai-draw-io Install dependencies: npm install Copy the env file and add your provider and API key, then start the dev server and open localhost:6002: cp env.example .env.local
Where can I get next-ai-draw-io?
The source code and installation instructions are on GitHub: https://github.com/DayuanJiang/next-ai-draw-io

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects