Next AI Draw.io

Next AI Draw.io turns written descriptions and uploaded files into diagrams you can edit on a draw.io canvas. It's for people creating technical diagrams who want AI help with drafting and revisions while keeping control of individual shapes.

• Diagram creation: Generate architecture diagrams, flowcharts and sequence diagrams, with AWS, Azure, GCP and Kubernetes icon libraries.
• Files to diagrams: Recreate a screenshot or sketch, or turn the contents of PDFs, Markdown and code files into a diagram.
• Chat and manual editing: Ask for changes to the whole canvas or selected shapes, then adjust labels, layout and styling by hand; each message includes your current canvas and manual edits.
• Versions and exports: Compare or restore AI revisions, undo changes, and export editable .drawio files or PNG and SVG images.
• Coding agent integration: An MCP server lets Claude Code, Cursor, Codex and VS Code create diagrams from local files and code, with a live canvas preview.

The app is open source under Apache 2.0. You can use the hosted web app, run it on your own server with Docker, or use desktop apps for Windows, macOS and Linux. It supports offline and intranet deployments, local models through Ollama, and OpenAI-compatible endpoints such as LM Studio and vLLM.

Cloud providers include OpenAI, Anthropic and Google AI; using them sends diagram context to the selected provider. Your own API key stays in your browser rather than in server storage. Model choice affects results: small models can produce invalid diagrams.