Have a feature request? Open an issue and let us know!
Most data tools require a backend, cloud account, or sending your data somewhere. DataStudio runs entirely in your browser. Python and DuckDB execute via WebAssembly, and files are stored locally using the browser's Origin Private File System. There is no server, no upload, and nothing to install.
This makes it ideal for:
- Quick data exploration: drag in a CSV or Parquet file and start querying in seconds
- Sensitive data: financial, medical, or proprietary datasets that can't leave your machine
- Offline work: everything works without an internet connection after the initial load
- Sharing analyses: notebooks use the standard Jupyter
.ipynbformat
| Feature | Description |
|---|---|
| Notebook interface | Multi-cell notebooks with Python, SQL, and Markdown cells. SQL cells automatically materialize as views accessible from Python. |
| In-browser execution | Python (Pyodide) and DuckDB run entirely in WebAssembly. Nothing leaves your machine. |
| Data visualization | Built-in charting (bar, line, area, scatter, pie) powered by ECharts. |
| File management | Upload, browse, and query CSV, Parquet, JSON, and Excel files. |
| Data validation | Built-in checks for uniqueness, not-null, accepted values, and custom SQL validations. |
| Excel editing with SQL | Query and transform Excel files directly using SQL. |
Open http://localhost:3000.
cp .env.example .env
docker compose up --buildContributions are welcome! Feel free to open an issue or submit a pull request.
MIT. Copyright (c) 2026 Dataspren

