Datasette Lite Now Runs Full Python Server App in the Browser with Service Workers and Pyodide
Datasette Lite has been rewritten to run the complete Datasette Python web application directly in the browser by leveraging WebAssembly, Pyodide, and a new service worker-based architecture. Developers can now launch configurable, isolated Datasette environments client-side for browsing SQLite, CSV, JSON, or SQL files, testing different versions, and demonstrating plugins with shareable URLs.
What changed?
Datasette Lite, a browser-based port of the Datasette Python application, has been completely rewritten to use service workers and Pyodide for running the full Datasette server app in the browser. This new implementation allows features previously limited to server deployments, including plugin support and asset loading, to work entirely client-side. Developers can configure Datasette Lite via URL parameters to load specific database files, select versions of Datasette or SQLite, and install plugins. Each browser tab runs an isolated, dedicated instance of Datasette Lite, providing independent environments.

Why does it matter to an everyday developer?
This update enables Python web application workflows that were previously not possible in the browser. Developers can now share links to live, browser-based Datasette environments showcasing specific features, data sets, or bug reproductions without launching a server or managing deployments. Plugin installation—including those with JavaScript or CSS assets—is fully functional. Tab isolation allows for simultaneous sessions with different configurations. The ability to select SQLite and Datasette versions is valuable for testing features and debugging across environments. Data and metadata files must only be publicly hosted with open CORS headers.
What can the developer do now?
- Use https://lite.datasette.io/ to launch a browser-only Datasette environment.
- Configure environments by URL parameters: load databases, install plugins, set metadata, or select Datasette and SQLite versions.
