Stop Manual Drag‑And‑Drop: The 34K‑Star AI Diagram Tool That’s Taking Over

Next AI Draw.io is an open‑source project that embeds an AI chat panel into draw.io, allowing users to describe a diagram in natural language and receive a fully editable draw.io chart. After generation, text, positions, colors, and connections can be further refined manually or via additional AI prompts.

Key Features

Text‑to‑Diagram Generation

The README provides a prompt example for a Retrieval‑Augmented Generation (RAG) architecture diagram. The AI produces nodes for data ingestion, retrieval, and generation, automatically arranging connections according to data flow, eliminating the need for templates or a blank canvas.

Dynamic connectors are also supported, making data‑flow visualizations clearer than static arrows.

Iterative Editing

After the first version is generated, users can request incremental changes such as adding approval nodes, merging modules, converting a flowchart to a swim‑lane diagram, or adjusting colors. The AI updates the existing diagram instead of recreating it from scratch. The author recommends submitting a few changes at a time, fixing structure first before tweaking layout and colors.

Upload of PDFs, Images, and Text

Beyond plain text, users can upload PDFs, text files, or images. PDFs and text files help organize diagrams, while images can be used to reconstruct legacy diagrams into editable versions, provided a vision‑capable model is selected. Sensitive internal documents should be processed on a self‑hosted instance.

Diagram Version History

Next AI Draw.io automatically saves diagram versions, allowing users to view or restore previous states, facilitating safe iterative refinement.

Cloud Architecture Diagram Support

The project includes icon libraries for AWS, GCP, and Azure. Generated drafts incorporate appropriate service icons, network boundaries, and call directions, producing diagrams that closely match common vendor‑specific conventions. Users must still verify permissions, network zones, and data flows before finalizing.

MCP Server Integration

The MCP Server exposes drawing capabilities to tools that support MCP, such as Claude Code, Cursor, and VS Code. These integrations can create new diagrams, read the current diagram, edit nodes, and export to draw.io, PNG, or SVG. Real‑time preview updates in the browser avoid manual re‑import steps.

When using MCP on a workstation with sensitive data, the local service should be isolated to prevent exposure to untrusted tools.

Model Service Configuration

The project does not lock users into a specific model. It supports providers like Doubao, DeepSeek, Kimi, Tongyi Qianwen, Zhipu, and local Ollama models. For simple flowcharts, a stable text model suffices; for image‑based inputs, a vision‑capable model is required. The documentation warns that overly small local models may only output reasoning without generating graphics.

Usage Options

Online Demo

Visit https://next-ai-drawio.jiang.jp/, open the chat panel, configure the model service and API key (stored locally in the browser), and try the tool without installation.

Docker Deployment

docker run -d -p 3000:3000 \
  -e AI_PROVIDER=kimi \
  -e AI_MODEL=kimi-latest \
  -e KIMI_API_KEY=YOUR_KEY \
  ghcr.io/dayuanjiang/next-ai-draw-io:latest

Access the service at http://localhost:3000. Environment variables can be managed via an .env file and the --env-file flag.

Running from Source

git clone https://github.com/DayuanJiang/next-ai-draw-io
cd next-ai-draw-io
npm install
cp env.example .env.local
npm run dev

Open http://localhost:6002. The .env.local file must contain model service and API key; without them the UI loads but diagram generation fails.

Desktop Application

Pre‑built binaries for Windows, macOS, and Linux are available on the Releases page. macOS builds lack code signing, so users may need to bypass the security warning as described in the release notes.

Suitable Scenarios

Drafting technical architecture diagrams by describing system modules, data flows, and exception branches.

Converting scattered textual process descriptions into structured flowcharts for review.

Reconstructing legacy diagrams from screenshots by uploading the image and generating an editable version.

Creating cloud‑vendor‑specific architecture diagrams with built‑in icon libraries for AWS, GCP, and Azure.

Generating quick diagrams within AI‑assisted coding environments (Claude Code, Cursor, VS Code) via MCP.

Author’s Perspective

The author acknowledges that AI‑generated images are no longer novel, but values Next AI Draw.io because the output remains editable in draw.io, avoiding full regeneration for each tweak. Nevertheless, the author stresses manual verification of business logic, permissions, and data flows, as the AI can produce incorrect branches without warnings.

A conservative workflow—accepting an initial structural draft then manually refining nodes and connections—maximizes productivity while ensuring correctness.

Open‑Source Repository

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