Bank import

Load a CSV statement and map its columns onto transaction fields.

You load a CSV file downloaded from your online banking and the app turns it into transactions in your budget.

The file is read in your browser

Same as receipts and credit reports. A bank statement is a complete record of where you have been and what you bought — the parsing happens on your device. Parsed operations reach the server when you request a match check.

Mapping the columns

There is no single CSV format, because every bank exports differently: another column order, another separator, another way of writing dates and amounts. So after loading the file you point out which column is the date, which the amount and which the description.

The app tries to work this out itself from the headers and the first few rows. The guess has to be confirmed — a misread amount column would turn the statement into a set of arbitrary numbers.

Preview and matching

Choose an account, a category and — if supplied by your bank — the transaction identifier column. Select Check for matches to send the parsed operations to the server for comparison with your budget. This does not save them yet; the CSV file itself stays on your device.

Each row explains why it will be added or skipped. Similar dates, amounts and descriptions on the same account require review: you may mark a row as a separate payment. Bank identifiers are compared within the selected account. An operation with the same identifier cannot be added twice.

Without identifiers, the number of occurrences matters. Two identical payments in one file remain two operations; importing that file again skips the existing counterparts. Do not mark a row as a separate payment when it is the same operation being imported again.

After reviewing, choose Import. Changing the account or column mapping requires another check. Data row numbers refer to parsed CSV entries, excluding the header.

History and undo

Below the form are the last 50 import batches. Undoing a batch removes only its unchanged transactions without linked documents. Later edits and payments linked to documents such as receipts are kept; the result shows their count. Operations saved before import history was introduced cannot be undone as a batch.