From a pile of documents to clean rows in five steps.
Upload one sample. The AI builds a Parser Template that reads every version of that document. Test it on your sample, choose where the data goes, then extract a whole batch at once.
Drop in one sample document
Upload one example. The AI reads it and builds the Parser Template in about 20 seconds.
One Parser Template per document type. Every design, every company.
A Parser Template is the list of details you want from a type of document: an order number, a date, a total, every line item. ParserBee reads each document by meaning, not by where things sit on the page, so one template covers every customer's or supplier's layout, as a PDF, a scan or a phone photo.
Start from your own sample, or from a ready-made template.
Most people start with the AI template builder: upload one typical document and the fields are ready in about 20 seconds. If your document is a common one, pick it from the library instead.
- Build it with AI from one PDF, PNG, JPG or WebP sample
- Or start from 83 ready-made templates across 16 industries
- Either way, review, edit or remove any field before you parse
Browse the document parser templates for invoices, orders, statements, logistics papers and more.
What goes in, what comes out.
Paper scans, supplier PDFs, photos from a phone, even an order that arrived as an email. Whatever goes in, it comes out as the same named columns.
Documents in
Up to 50 MB a file. PDFs up to 24 pages on paid plans, 12 on the free trial.
Data out
QuickBooks, Xero and your ERP are reached through Zapier, Make, n8n, a CSV import or the API.
Four ways to run it, once the template is ready.
The same Parser Template works in all four, so you can start with a batch upload today and automate it next month.
Upload a batch in Data Lab
Drop in up to 100 files, press Extract Data and download one sheet. Best for month-end piles and backlogs.
How Data Lab works →Fill a Google Sheet or Airtable table
Link a sheet or table to your Parser Template. Every document you extract from then on is added as a row.
PDF to Google Sheets →Catch documents as they arrive
Zapier, Make or n8n pick up attachments from an inbox or files from a Drive folder and send them to ParserBee.
See all integrations →Call the API from your own software
One request with the file and your Template ID returns the fields as JSON. Python and Node.js libraries included.
Developer docs →Every batch, one click away.
Each Data Lab batch is listed in Activity with its Parser Template, the number of documents and the result. Open a batch to see the same spreadsheet and export it again as Excel, CSV, PDF or Google Sheets.
- Filter by Parser Template and date
- See at a glance how many documents succeeded or failed
- Your remaining page credits are always on screen
From messy documents to clean rows.
One page credit per page. That is the whole model.
How ParserBee works: questions
What is a Parser Template?
A Parser Template is the list of details you want from one type of document, such as an order number, a date, a total and every line item. Each field has a name, a type (Text, Number, Yes / No, Group or List) and a short plain-English description that tells the AI what to look for.
Do I need a separate template for every supplier or customer?
No. You need one Parser Template per document type, not per company. ParserBee reads each document by meaning, not by where a value sits on the page, so one invoice template reads every supplier's invoice, in any design, as a PDF, scan or photo.
How do I build my first Parser Template?
Two ways. Upload one sample document and the AI template builder designs the fields in about 20 seconds, or pick one of 83 ready-made document parser templates. Either way you review the fields and run a test on your own file before you parse anything else.
What file types can ParserBee read?
PDFs, scans and phone photos (JPG, PNG, WebP and HEIC), plus Excel (XLSX), CSV, email (EML), PowerPoint, plain text and HTML files. DOCX and the old XLS format are not accepted: save them as PDF first. Files can be up to 50 MB, and PDFs up to 24 pages on paid plans (12 on the free trial).
How long does it take to get my first rows?
About 2 minutes. The AI builds a Parser Template in about 20 seconds, and the test run on your sample shows the rows straight away.
Can I change the fields later?
Yes. Open the Parser Template and edit it: rename a field, add one, remove one, or sharpen its description. Changes apply to the next document you extract. You do not need any technical skills.
How do documents get into ParserBee without uploading them?
Do I need a developer?
No. Building a template, testing it, bulk extraction in Data Lab and the Google Sheets and Airtable links all work in the browser with no code. The API is there if you want it.
How much does it cost?
New accounts get 25 free pages, no credit card. After that, plans start at $29 a month for 250 pages; see pricing. One page of a document uses one page credit, and the test run counts like any other extraction.
Is my data used to train AI?
No. Your documents are never used to train AI models, and files travel to ParserBee over an encrypted (HTTPS) connection.