SQLite · Playground

Run SQL on a CSV without uploading it

1. Drop a CSV, TSV or JSON file onto the box above, or click to choose one. Each file becomes a table, with columns taken from the header row (CSV) or the object keys (JSON).
2. Write SQL in the editor and press Run (or ⌘/Ctrl + Enter). Click a table name to drop a SELECT into the editor.
3. Read the results in the grid, then copy them or export as CSV or JSON. Nothing is uploaded at any point.

This is real SQLite, not a toy parser

The engine here is the genuine SQLite C library compiled to WebAssembly via sql.js, running inside your browser tab. That means the full SQLite dialect works: JOIN, GROUP BY, HAVING, subqueries, common table expressions (WITH), window functions, CASE, the built-in string/date/math functions, and CREATE/INSERT/UPDATE/DELETE. You can load several files and join across them, or ignore files entirely and use it as a scratch database by typing CREATE TABLE and INSERT yourself.

Analyse a spreadsheet with SQL

A CSV export from a spreadsheet, a database dump, or an API response saved as JSON — drop it in and query it the way you would a real table. Numbers in a CSV are detected and stored as numbers, so SUM, AVG, MIN/MAX and comparisons behave as expected. Load two files and JOIN them on a shared key to combine data sets locally, with no database server to set up.

Is my data private?

Yes. The SQLite engine and your files stay in the browser tab — there is no upload step, no account, and no analytics. You can open your browser's network tab and confirm that querying sends nothing. Once the page has loaded, the whole tool works offline. The in-memory database is discarded when you close or reload the tab, so nothing is persisted either.

How much data can it handle?

Everything runs in memory in a single tab, so it comfortably handles files up to tens of megabytes — plenty for most CSV exports and log dumps. Very large files (hundreds of MB) can exhaust the tab's memory; for those, a desktop SQLite is a better fit. The results grid shows the first 1,000 rows for speed, while exports include every row your query returned.