Intelligent document processing without a platform project
Most finance and operations teams do not need a long platform rollout to stop retyping documents. They need clean fields from invoices, forms and statements, a way to check the uncertain ones, and a predictable bill. PdfParse gives you that as one API and a web app, priced per page.
Check one of your own documents first
Before any budget conversation, drop a real invoice or form and see what comes back. Type the fields your team keys today into "Fields to extract" and compare.
What one call returns
Page one rendered with every block outlined, next to the JSON your code receives.
- Reading text layer
- Detecting layout
- Extracting tables
- Building JSON
Headings, paragraphs, tables and totals detected
Portsmouth, NH 03801
| Description | Qty | Unit | Amount |
|---|---|---|---|
| Oak countertop, 2.4 m | 2 | 640.00 | 1,280.00 |
| Cabinet doors, shaker | 12 | 85.00 | 1,020.00 |
| On site fitting | 1 | 450.00 | 450.00 |
Quote the invoice number with your transfer.
Goods remain ours until paid in full.
Pages 2 and later are parsed after you create an account.
{ "document": { "pages": 2, "parsed_pages": 1, "language": "en" }, "pages": [ { "number": 1, "blocks": [ { "type": "heading", "text": "Invoice", "bbox": [48, 52, 132, 84] }, { "type": "paragraph", "text": "Harbor Street Joinery, 14 Dock Lane, Portsmouth" }, { "type": "key_value", "pairs": { "Invoice no.": "INV-2026-0418", "Issued": "Oct 14, 2026", "Due": "Nov 13, 2026" } }, { "type": "table", "columns": ["Description", "Qty", "Unit", "Amount"], "rows": [ ["Oak countertop, 2.4 m", 2, 640.00, 1280.00], ["Cabinet doors, shaker", 12, 85.00, 1020.00], ["On site fitting", 1, 450.00, 450.00] ] }, { "type": "total", "label": "Total due", "value": 2983.75, "currency": "USD", "confidence": 0.98 } ] } ], "fields": { "invoice_number": { "value": "INV-2026-0418", "confidence": 0.99, "page": 1 }, "total": { "value": 2983.75, "confidence": 0.98, "page": 1 }, "due_date": { "value": "2026-11-13", "confidence": 0.97, "page": 1 } } }
Downloads from the demo include a watermark field and cover page one only.
Page one is parsed in your browser, so your file never reaches our servers.
One demo per day from this network. Create an account to parse more pages.
$ curl https://api.pdfparse.ai/v1/parse \ -H "Authorization: Bearer $PDFPARSE_API_KEY" \ -F file[email protected] \ -F outputs=json,markdown,csv
Where the hours go today
In a typical back office, documents arrive by email, portal download and scanner. Someone opens each file, finds the supplier, the date, the totals and the line items, and types them into the accounting or operations system. A second person spot checks a share of the entries. Errors surface weeks later during reconciliation.
The work is repetitive, but the documents are not identical. Layouts change between suppliers, scans arrive skewed, and tables run across pages. That variety is the reason simple templates break and the reason manual keying survives.
Three ways to handle it
| Approach | What it takes | What you pay for |
|---|---|---|
| Manual keying | Staff time for every document, plus checking | Hours, rework and late corrections |
| Platform project | Vendor selection, integration services, template training, licence negotiation | A long project before the first document is processed |
| One API priced per page | An API key, a schema per document type, a review step for uncertain fields | A monthly plan with a fixed page allowance |
PdfParse sits in the third row. Your developers connect it once. Your operations staff keep control of what is extracted through schemas they can read and edit.
The workflow, step by step
-
1
Intake. Files arrive through the API from your own systems, or staff drop them into the web app. Batches of up to 500 files per job suit month end volume on Growth and above.
-
2
Extraction. Each document is parsed for layout, tables and text, with OCR on scanned pages. Your schema defines the fields, such as supplier name, invoice number, due date, net amount and line items.
-
3
Confidence routing. Every field returns with a confidence value and the page it came from. You decide the threshold. Fields above it flow straight into your system, fields below it go to a person.
-
4
Review. Reviewers see the low confidence field next to the highlighted area on the page, correct it if needed and move on. They check a few fields instead of keying whole documents.
-
5
Export. Approved data leaves as CSV for import, as JSON to your services, or through a webhook the moment a job finishes.
Why review of low confidence fields matters
A useful document workflow does not pretend every value is certain. Faded scans, handwritten amendments and unusual layouts happen. PdfParse reports how sure it is about each field rather than one score for the whole file, and it tells you which page the value came from and where on that page.
That changes the shape of the job. Your team stops being a typing pool and becomes a review desk. The volume they touch shrinks to the fields that deserve a human look, and the audit trail shows what was corrected.
Cost you can forecast
Every plan has a monthly page allowance. Starter covers 2,000 pages for $49 a month, Growth 7,000 pages for $149, Scale 24,000 pages for $499 and Enterprise 72,000 pages for $1,499. Paying yearly halves the monthly price. Custom schemas, which this workflow relies on, start on Growth. There are no per field fees, no template charges and no services contract.
If you reach the allowance, parsing pauses with an Upgrade button and nothing is charged automatically. Finance never sees a surprise overage. Full plan details are on PDF extraction API pricing.
Controls your security review will ask about
Files are encrypted in transit and at rest. Retention follows the plan, from 7 days on Starter, to a setting up to 30 days on Growth, to a zero retention mode on Scale and Enterprise. Scale adds an audit log and usage reports per API key. Enterprise adds SSO with SAML, roles and permissions, a DPA, priority support, EU or US processing region and invoicing with net terms after the first paid month. Read the details on secure document processing.
A realistic first month
Week one, pick the document type with the most volume and build its schema in the web app. Week two, run a backlog of real files and set the confidence threshold with the people who key them today. Week three, connect the webhook or CSV export to your system. Week four, measure how many fields still need review and adjust the schema descriptions where values were missed.
Your developers will find request and response formats in the document parsing API reference. The broader module list is on the data extraction tool page, and the model side of the engine is explained under AI document processing.
FAQ
Do we need a developer to start?
Not for a first evaluation. The web app lets operations staff upload files, build a schema and export CSV. A developer is useful when you connect PdfParse to your accounting or operations system through the API or webhooks.
How does the review step work?
Every extracted field has a confidence value and a source page. You set the threshold. Fields below it are shown to a reviewer next to the highlighted area of the page, and confirmed values are exported like the rest.
What document types can we process?
Anything that arrives as a PDF, scanned PDF, PNG, JPG or TIFF. Typical starting points are supplier invoices, receipts, application forms, bank statements and contracts. Each type gets its own schema.
How is this priced compared with a platform licence?
PdfParse is a monthly subscription with a page allowance, from $49 a month for 2,000 pages up to $1,499 a month for 72,000 pages, with half price on yearly billing. There are no setup or services fees.
Can we keep documents in the EU?
Enterprise lets you choose an EU or US processing region and includes a DPA. Zero retention mode on Scale and Enterprise removes files and results once they are delivered.
Replace keying with review this month
Start with one document type and one plan. Upgrade when volume grows.