ChartDB
AGPL-3.0
- Platform
- Web Docker
- Licence
- AGPL-3.0 Open source
ChartDB is an open-source, browser-based editor for database diagrams. You paste one “smart query” against your database, it returns the schema as JSON, and ChartDB draws the entity-relationship diagram from it — no install, no connection string, and no database password handed to the tool.
From there you edit the diagram like a whiteboard: add and rename tables and columns, annotate, drag relationships into something readable, and export the result. The export can target a different dialect from the source, so a MySQL schema can be turned into PostgreSQL DDL, which makes it useful for migrations and for documenting a schema nobody wrote down.
Features
- Instant schema import from a single query, with no credentials required
- Interactive diagram editor for tables, columns, keys and relationships, with annotations
- DDL export in the dialect of your choice, including AI-assisted conversion for migrations
- Supported sources: PostgreSQL (including Supabase and Timescale), MySQL, MariaDB, SQL Server, SQLite (including Cloudflare D1), CockroachDB and ClickHouse
- Runs as a static web app, an npm dev server, or a Docker container, and there is a hosted version at app.chartdb.io
Things to know
- The project still calls itself Public Beta.
- The AI-assisted DDL export needs an OpenAI API key supplied at build or container start; without one the diagram and SQL export still work.
- The import depends on running the generated metadata query yourself against the target database, so you need read access and a client to run it in.
Alternatives
drawDB is the other well-known open-source database diagram editor and needs no metadata query at all — you describe the schema in its own editor. ChartDB’s advantage is going straight from a live database to a diagram.
Views: 2