Is my CSV uploaded to a server?

No. editsqlite is a static site with no server behind it. Your CSV is read in the browser tab, the table is built in memory, and the .sqlite lands in your downloads folder. Turn off your network and it still works.

How are column types decided?

The import wizard reads the whole column before deciding: if every non-empty value is a whole number you get INTEGER, a decimal anywhere gives REAL, and one stray word makes it TEXT. The guess is only a default — each column has a dropdown you can override, and a value that can't fit the chosen type is kept as text rather than mangled.

Does it handle commas and line breaks inside fields?

Yes. The parser follows RFC-4180 quoting, so quoted fields containing commas, line breaks and doubled quotes parse correctly. CRLF and bare CR both end a row.

Is the resulting .sqlite a real SQLite file?

Yes — it is written byte by byte by our own SQLite writer and passes the official engine's PRAGMA integrity_check. You can open it in any SQLite tool, and the engine will insert its own rows into it.

Converters · CSV → SQLite

Convert CSV to SQLite — in your browser, nothing uploaded

A CSV is a flat sheet of text; a SQLite database is a typed, queryable file. The gap between them is usually a Python script or a command-line import that makes you declare every column by hand. Here it's a two-minute job in a browser tab: open the CSV, glance at the types the wizard picked, and download a real .sqlite file. Your data never leaves the machine.

the CSV goes in as text; a typed, certified database comes out.

How to convert a CSV to SQLite

Open your CSV. On the editor, click Import CSV — or just drop the .csv onto the page. You can also paste rows straight into the wizard.
Check the column types. The wizard surveys every column and proposes INTEGER, REAL or TEXT. Each has a dropdown — override any guess before the table is built.
Create the table. Name it and create it. It opens in the grid, where you can fix a cell, add a column or run a quick SELECT before exporting.
Export the .sqlite. Click ⬇ .sqlite (certified). A fresh database file is assembled in the tab and downloaded.

Convert a CSV now →

Why the column-typing step matters

Most CSV importers either make everything text or guess from the first row and get it wrong the moment row 900 breaks the pattern. editsqlite reads the whole column before deciding: every non-empty value has to agree. All whole numbers give you INTEGER; a single decimal anywhere makes it REAL; one stray word makes it TEXT. The guess is a starting point, not a verdict — you set the final type yourself, and a value that can't honour it is kept as text instead of being mangled, the same forgiveness SQLite itself shows.

Good to know

Quoted fields with commas, line breaks and doubled quotes parse correctly (RFC-4180).
Empty cells become NULL, not empty strings — so IS NULL works as you'd expect afterward.
Very wide single rows (over ~4KB) aren't supported yet; ordinary spreadsheet rows are fine.
The finished database is a normal SQLite file — open it anywhere, and it passes PRAGMA integrity_check.

Other conversions

SQLite → CSVThe round trip: open a .sqlite and export a table back to CSV. MySQL → SQLiteOpen a mysqldump .sql file and leave with a working .sqlite. All convertersCSV, MySQL, Postgres and SQLite — every direction the tool handles.