Convert CSV to SQLite

How to convert CSV to SQLite

1. Open a .csv, .tsv or .txt file, drop it on the tool, or paste rows. Check the table under Input; set the delimiter, header row or encoding under Import settings if needed.
2. Check the Table name. It starts as the file name with spaces and punctuation turned into underscores, so orders 2026.csv becomes orders_2026.
3. Choose Value types, an optional Primary key, and whether to Write empty cells as NULL.
4. Select Download SQLite, then open the file, for example with sqlite3 orders.sqlite and SELECT * FROM orders LIMIT 10;.

The first conversion downloads the SQLite engine (sql.js, compiled to WebAssembly) once from assets.viewcsv.app. If that download fails, the tool shows an error and you can try again; your file stays on this device.

To convert many files with the same settings, choose Batch: convert several files. Each file becomes its own database, with a table named after that file, and they download together as one ZIP. In a batch, each file must be 25 MB or smaller.

What the database contains

One table, created with CREATE TABLE "name" (…), with one column per CSV column in the original order and all rows inserted in a single transaction. There are no extra indexes, views or ID columns.

Line breaks, quotes and commas inside quoted CSV values are stored inside the value unchanged.

Troubleshooting

• The conversion stops at the primary key. The chosen column has an empty or repeated value. Find duplicates with Validate CSV or the CSV viewer, choose another column, or set Primary key to None.
• The engine could not be downloaded. The SQLite WebAssembly file comes from assets.viewcsv.app. Check your connection, firewall or content blocker, then try again.
• `SUM()` treats a column as text. The column is TEXT. Convert again with Detect numbers and true/false; if it still stays text, one value isn't a plain number (1,234, 12 kg, 007). Or cast it in SQL: CAST(total AS REAL).
• The file is too large for this tool. SQLite conversion works on files up to 25 MB, 100,000 rows, 250 columns and 2 million cells. For a bigger file, generate a script with CSV to SQL, which streams large files, and run it with sqlite3.

Limits and privacy

The CSV must open in the browser workspace: up to 25 MB, 100,000 rows, 250 columns and 2 million cells. The database is built in memory and exported as one file.

The engine runs in a background worker in this tab, so the page stays responsive and Cancel stops it. Only the engine file is downloaded; the CSV and the database never leave your device.

Example

An orders file with leading-zero IDs, an empty date and a two-line note, converted with Detect numbers and true/false, table name orders and order_id as the Primary key, then opened with the sqlite3 shell. total became REAL and paid became INTEGER 1/0, while order_id and shipped_on stayed TEXT. With the default Keep every value as text, all six columns are TEXT and total keeps 129.50.

order_id,customer,total,paid,shipped_on,note
000417,"Okafor, Ada",129.50,true,2026-04-02,
000418,Mei Tan,15,false,,"Leave at door
side gate"

sqlite> .schema
CREATE TABLE "orders" ("order_id" TEXT PRIMARY KEY, "customer" TEXT, "total" REAL, "paid" INTEGER, "shipped_on" TEXT, "note" TEXT);

sqlite> SELECT order_id, total, typeof(total), paid, shipped_on, quote(note) FROM orders;
000417|129.5|real|1|2026-04-02|NULL
000418|15.0|real|0||'Leave at door
side gate'