Convert a bank statement to CSV, free

Two CSV downloads, for two jobs

Most imports reject anything above the header row, so choose the transactions-only file when the CSV is going into software. The file starts with a UTF-8 byte-order mark, so Excel opens Urdu, Arabic and accented descriptions correctly. Values that a spreadsheet could run as a formula are prefixed with an apostrophe, so a hostile description can't execute.

Import the CSV into accounting software

Each accounting package wants its own columns, so expect a short mapping step on import. The transactions file uses the column names printed on your statement.

• QuickBooks Online accepts a three-column file (Date, Description, Amount) or a four-column one (Date, Description, Credit, Debit). Delete the Balance column and any extra columns before you upload.
• Xero's statement import maps Date and Amount, plus optional Payee, Description and Reference. One signed Amount column is what it expects.
• Wave, Zoho Books, Sage and FreshBooks have similar importers. Check the date format your software expects, and set it when it asks.
• Google Sheets: File, Import, Upload, then choose comma as the separator.

Bank statement to text, CSV or JSON

If you searched for bank statement to text, a CSV is usually what you want: it is plain text that any program can open, with the structure kept. For code, download JSON instead, which keeps each field's name and each table's columns. Raw OCR text, by contrast, loses which number belongs to which column.

Common problems and fixes

• Password-protected PDF: open it, print or save it without the password, then upload the copy.
• Amounts with Dr and Cr: they're kept as printed. Add a column that makes Dr negative before importing.
• Commas in amounts such as 2,400.00: they're quoted in the CSV, so the column stays intact. If your importer wants plain numbers, remove thousands separators with find and replace.
• Only part of the statement came through: the free try reads page 1. A free account reads every page and joins them into one table.

curl -X POST "$DOCREFINERY_API/v1/workflows/wf_demo_bank_statement/run" \
  -H "Authorization: Bearer $DOCREFINERY_API_KEY" \
  -F "[email protected]"

# Take trace_id from the answer, then poll until status is "fulfilled"
curl "$DOCREFINERY_API/v1/runway/9f1c0a7e" \
  -H "Authorization: Bearer $DOCREFINERY_API_KEY"

{
  "trace_id": "9f1c0a7e",
  "status": "fulfilled",
  "document_type": "bank_statement",
  "result": {
    "fields": [
      {
        "key": "client_name",
        "alias": "Client Name",
        "text": "Alex Sample",
        "field_confidence": 0.99,
        "location": { "page": 1, "bbox": [0.08, 0.12, 0.46, 0.14] }
      },
      {
        "key": "account_number",
        "alias": "Account Number",
        "text": "0098 7654 3210",
        "field_confidence": 0.98,
        "location": { "page": 1, "bbox": [0.08, 0.17, 0.46, 0.19] }
      },
      {
        "key": "statement_start_date",
        "alias": "Statement Start Date",
        "text": "2026-10-01",
        "field_confidence": 0.97,
        "location": { "page": 1, "bbox": [0.08, 0.22, 0.46, 0.24] }
      },
      {
        "key": "transactions",
        "alias": "Transactions",
        "table": [
          [
            { "key": "date", "alias": "Date", "text": "2026-10-01" },
            { "key": "description", "alias": "Description", "text": "Transfer from current account" },
            { "key": "debit", "alias": "Debit", "text": "" },
            { "key": "credit", "alias": "Credit", "text": "500.00" },
            { "key": "balance", "alias": "Balance", "text": "13,000.00" }
          ]
        ]
      }
    ],
    "extraction_approval": "auto",
    "low_confidence_fields": []
  }
}