Database Schema Visualizer

Unsupported input

The tool may reject input that does not match the expected content, structure, or file type.

Fix: Confirm the tool input requirements and paste the correct type of data.

Incomplete values

Missing fields or partial content can block processing or produce weak results.

Fix: Provide the full required input before running the tool.

Copying placeholder content

Sample or placeholder values can lead to output that looks valid but is not ready for real use.

Fix: Replace placeholders with your actual values before relying on the result.

How is this different from dbdiagram.io?

dbdiagram.io is a full editor with collaboration and project management. This tool is a one-shot viewer — paste a schema, get an ERD, export, move on. We support importing the same DBML format so you can move between the two without rewriting.

Which SQL dialects are supported?

ANSI SQL `CREATE TABLE` syntax with vendor extensions for PostgreSQL, MySQL, and SQLite. The parser is intentionally lenient — it skips unrecognised clauses rather than failing, so a `pg_dump --schema-only` output usually visualizes without edits.

Can it parse a multi-statement migration file?

Yes. The parser walks the input statement-by-statement and groups all `CREATE TABLE` plus `ALTER TABLE ... ADD CONSTRAINT FOREIGN KEY` into a single schema model. Standalone indexes show up in the per-table panel.

How big a schema can it render?

Comfortably up to ~100 tables. Beyond that dagre layout slows down and the canvas becomes hard to navigate; consider splitting the schema by domain.

Does it support views, materialized views, or stored procedures?

Not yet. The viewer is scoped to tables and FK relationships. Views and materialized views are detected and listed in a small sidebar but not laid out on the canvas.

Is my schema sent to a server?

No. SQL parsing, DBML parsing, and rendering are all client-side. Nothing leaves your browser.