Bank statement converter
Statements in. Transactions out.
Upload a statement as a PDF or a scan, forward it by e-mail or send it to the API. Parsevix returns the account, the balances and every transaction as structured data, ready to reconcile or import.
- Every transaction with date, amount and balance
- Names and IBANs masked before the AI step
- Online-banking PDFs and scans, in many languages
No credit card · 20 free pages every month
What you get back
The account, the balances and every transaction.
The built-in Bank Statement category returns 22 fields: the account, the statement, the balances, the period totals and the transactions as a table. Use it as it is, or change it in your project.
Dates come back as ISO dates and amounts as plain numbers, however the statement prints them. A field the statement doesn't show, such as a reference on a card payment, is left out, not guessed.
Account
account_infoobject
- bank_namestring
- account_numberstring
- account_holderstring
- ibanstring
- swiftstring
- account_typestring
- currencystring
- branchstring
Statement
- statement_periodstring
- statement_numberstring
- generated_datedate
Balances
balancesobject
- opening_balancenumber
- closing_balancenumber
Period totals
summaryobject
- total_creditsnumber
- total_debitsnumber
- net_changenumber
Transactions
transactionsarray
- tx_datedate
- tx_descriptionstring
- tx_amountnumber
- tx_typestring
- running_balancenumber
- referencestring
Your own fields
A name, a type and one sentence on what it is.
- value_date
- counterparty_iban
- end_to_end_id
- overdraft_limit
Made for statements
Long tables, private data, many banks.
Statements are long tables full of names and account numbers, and every bank lays them out its own way. This is what Parsevix does with them.
What the AI model sees
A bank statement is mostly personal data: the account holder, the IBAN, the name behind every transfer. Parsevix finds them and replaces them with placeholders before any text goes to the AI model, then puts the real values back into your result.
Detection is automatic: it catches most personal data, but not necessarily every item. If masking fails, the document stops there.
- 02
Several pages, one list
A statement of up to 30 pages is read as one document: the transactions of every page come back in one list, with the balances from its first and last page.
- 03
Balances to check against
Opening and closing balance, total credits and total debits are fields of their own, so your import can check that the transactions add up before it books anything.
- 04
Every column of the table
Each transaction has its date, description, amount, credit or debit, running balance and reference, as the statement prints them.
- 05
Online-banking PDFs and scans
A PDF downloaded from online banking and a scanned paper statement are read the same way: every page is read as an image, and sideways pages are turned upright first.
- 06
Any bank, any language
The language is detected on every page. A statement from a bank in another country fills the same fields, with ISO dates and plain numbers.
- 07
Every value linked to its line
Each value is linked to the spot on the page it came from. Click it in the dashboard to see the line on the statement.
How statements get in
Upload them, forward them or send them from your system.
Dashboard
Upload statements as they come from online banking. Check each transaction next to its line on the page and approve the result.
E-mail address
Paid plansEvery project gets its own address. Forward the statements your bank e-mails you, and each attachment becomes a document.
API
Send statements from your own system and get the transactions in the same response, or receive results by signed webhook or through Zapier.
Request
curl https://parsevix.com/api/v1/extract \ -H "X-API-Key: $PARSEVIX_API_KEY" \ -F file=@statement.pdf \ -F category="Bank Statement"
The request stays open until the document is read and answers with the JSON.
- POST /documentsUpload without waiting; the result arrives by webhook or callback.
- POST /extract-autoNo category: every field Parsevix finds, named as in the document.
- POST /classifyOnly the document's category, for 1 credit.
Response · trimmed
Security & compliance
- PII maskedautomatically, before the AI step
- Encryptedin transit and at rest
- No AI trainingon your documents
- API files deletedafter processing
Priced per page, not per feature.
Every plan reads, sorts and masks documents the same way. Picking the category is included, and a document that fails costs nothing.
See pricing- credit per page
- 1
- free pages every month
- 20
- a month for 500 pages
- $39
Statement FAQ
Bank statements,
answered.
What teams ask before they send their first statement.
PDFs from online banking, scans and photos (JPG, PNG, TIFF) and Word, up to 10 MB and 30 pages each. A PDF that needs a password to open can't be read: the document stops with a message saying so and costs nothing. Remove the password and upload it again.
Not as a file: Parsevix returns the transactions as JSON, through the API, by webhook or in the dashboard. Zapier can pass them on to a spreadsheet such as Google Sheets, and in your own code a JSON list of transactions takes a few lines to turn into CSV.
A file can have up to 30 pages. Send a longer statement as several files, for example one per month; each comes back with its own transactions and balances.
Names and IBANs are masked automatically before any text goes to the AI model and put back in your result; detection catches most personal data, not necessarily every item. Statements uploaded in the dashboard or by e-mail stay in your project until you delete them, and files sent through the API are deleted after processing unless you turn on keeping them. Your statements are never used to train AI.
Yes. Change the Bank Statement category in your project and add the fields you need, such as the value date of each transaction, the counterparty's IBAN or an end-to-end reference, each with a sentence on what it is. A field is filled only from what the statement shows.
One credit per page: a four-page statement uses four. Choosing the category is included, and if a document fails, its credits are given back.
Related use cases