Receipt & invoice parsing APIPaper in.
JSON out.

Send a receipt, invoice or phone photo to one endpoint. Get back merchant, dates, totals, tax and line items, plus a confidence score that tells you when a person should take a look. Billed per document, with up to 10 pages in each.

✓ 20 docs/month free✓ no credit card✓ webhooks on Starter and Pro
Trace a field
FITZPATRICK AND SONS
00480 Cook Cove
Spencerport, UT 12036
INVOICE
No.12847181Date03/03/2012Bill toDuncan PLC
QTYDESCRIPTIONPRICE
4HP Desktop PC Core i5 16GB139.95
3Custom AMD Ryzen Threadripper1,400.00
1Dell Optiplex Dual Core 3.4GHz217.00
3Dell Optiplex 790 i7 Quad159.99
2Pentium 133MHz Desktop Tower390.00
Subtotal6,236.77
Tax 10%623.68
TOTAL USD$6,860.45
GET /documents/{id}● completed
{
"document_type": "invoice",
"invoice_id": "12847181",
"date": "2012-03-03",
"currency": "USD",
"merchant": "Fitzpatrick and Sons",
"customer": "Duncan PLC",
"subtotal": "6236.77",
"tax_rate": "10.00",
"tax": "623.68",
"total": "6860.45",
"line_items": [ …5 items ]
}

Sample: a real invoice and the fields DocuParse API returned for it.

Now try yours. Drop a PDF, JPG or PNG.no account · one document →
One price
Up to 10 pages per document
11–20 pages count as two. Documents that fail extraction don’t count at all.
4
formats, scans included
PDF, JPG, PNG and CSV, including phone photos and scans. Up to 10 pages per document.
$11.99
per 1,000 documents on Starter
No minimum, no annual contract, no sales call. Pro brings it to $10.00.

10 pages, one price

Every document includes up to 10 pages. A 10-page invoice counts as one document, the same as a one-page receipt.

On per-page APIs such as AWS Textract and Mindee, that 10-page invoice is 10 billable pages. On Docparser it's 2 credits. On DocuParse API it's one document.

Compare monthly costs →

§ 01 · How it works

Two calls. That’s the whole integration.

One bearer token, one submit endpoint. No OCR pipelines. No template setup. No prompt engineering.

01
POST /api/v1/extractSend a receipt or invoice as PDF, JPG, PNG, or CSV using a simple multipart POST request. You get a document_id back straight away.
02
Rules read, AI checksRule-based parsing reads every field, then an AI model checks the result and corrects what it finds wrong before you get it.
03
GET /api/v1/documents/{id}Get structured data including merchant, date, totals, tax and line items, with a confidence score and a needs_review flag.
04
or your webhookOn Starter and Pro, skip polling. Results delivered instantly to your webhook.
# 1. Submit the document
curl -X POST https://docuparseapi.com/api/v1/extract \
  -H "Authorization: Bearer $DOCUPARSE_API_KEY" \
  -F "file=@invoice.pdf"
# → {"success": true, "document_id": "<id>", "status": "queued"}

# 2. Fetch the result (repeat until "status" is "completed" or "failed")
curl https://docuparseapi.com/api/v1/documents/<id> \
  -H "Authorization: Bearer $DOCUPARSE_API_KEY"
Full guides for Python, Node.js and Next.js in the developer docs.

§ 02 · Honest by default

It tells you when it isn’t sure.

Rule-based parsing reads every field, then an AI model checks the result and corrects what it finds wrong before you get it. Every response carries confidence and needs_review, so clean documents flow straight through and messy ones go to a person first.

Totals reconcile, nothing flagged
"needs_review": false
→ Write it straight to your books.
Scanned invoice from Fitzpatrick and Sons
A field the checks couldn’t confirm
"needs_review": true
→ Route it to a person before it posts.

§ 03 · Who it’s for

Built for developers and product teams.

01SaaS founders

Building expense, accounting, or bookkeeping tools that need document ingestion.

02Developers

Skip OCR pipelines. Get extraction as an API call in your existing stack.

03Finance teams

Automate receipt processing and invoice data entry for your team.

04Agencies

Process batches of business documents for clients without custom infrastructure.

§ 04 · Every month, at any volume

For projects that handle receipts every month.

1,000 documents a month

205,000
120

Public list prices, verified 20 October 2026. DocuParse API: up to 10 pages per document; 11–20 pages count as two. Textract: AnalyzeExpense per page. Google: Invoice parser, per document up to 10 pages. Veryfi: free up to 100 documents a month, then the Starter invoice rate with its $500/month minimum. Sources →

DocuParse APIStarter plan, flat$11.99
AWS Textract$0.01 × 2 pages$20.00
Google Document AI$0.10 per 10 pages$100.00
Veryfi$500 minimum$500.00

§ 05 · What you get

Built for apps that process documents every month.

security
Secure API keys

Keys are hashed and stored securely. Raw key shown only once.

How keys are stored →
dashboard
Financial dashboard

See total spend, vendors, categories, and trends from your extracted documents.

Open a free account →
starter · pro
Webhook delivery

Results delivered instantly to your webhook.

Webhook docs →
snippets
Copy-paste snippets

cURL, Python, Node.js, and Next.js examples in the docs.

See code examples →
output
Consistent JSON

Normalized output for key fields like merchant, totals, tax, dates, and line items.

JSON schema →
errors
Clear error codes

Typed error responses so your error handling is predictable.

See error codes →

§ 06 · By document type

Explore by document type.

Purpose-built extraction for every business document.

§ 07 · Fits your stack

Already speaks your tools.

Use DocuParse API with no-code tools, developer runtimes, and finance systems.

All integration guides →