Send a receipt or invoice image, get back typed JSON — vendor, line items, totals, tax, dates — in a couple of seconds. Built on a vision LLM, so it reads crumpled phone photos without a training set. Priced for indie devs, not enterprise procurement.
No ML pipeline to build, no per-vendor templates, no rules engine. Post an image, get a typed object back.
Send a receipt or invoice as a file or base64 string. Phone photos, scans, and multi-page PDFs all work.
A vision model extracts vendor, dates, totals, tax, payment method, and every line item — even from messy layouts.
A stable, documented schema you can drop straight into your accounting, expense, or bookkeeping product.
$ curl -X POST https://api.receiptparse.dev/v1/parse \
-H "Authorization: Bearer $KEY" \
-F file=@receipt.jpg
{
"doc_type": "receipt",
"vendor": "GreenLeaf Market",
"date": "2026-09-06",
"currency": "USD",
"subtotal": 31.95,
"tax": 0.00,
"total": 31.95,
"payment_method": "VISA ****4471",
"line_items": [ /* … */ ]
}
Enterprise OCR charges enterprise rates for volume you don't have. Swap it out for a flat monthly plan.
We're onboarding early users now. Drop your email and we'll send an API key with free pages to build against — and a launch discount when billing goes live.