Pricing
1¢ per document
Same price for everyone — from 1 to 100 million documents.
You pay only for documents we actually recognise.
Blurry photo, no document in frame, unknown type — free. You never pay for a failed scan.
How billing works
A call is charged when the document type has been determined and data was actually extracted from it: a machine-readable zone whose check digits pass, at least five fields of the visual zone, or a correctly decoded barcode. Nothing else is charged.
No document found, an unreadable image, a type we do not support, an internal error, a timeout — none of these cost anything.
Every response carries a billed flag, so your own records match ours without a reconciliation step.
| Result | What it means | Cost |
|---|---|---|
| Recognized | The type was determined and the data came out. This is the only outcome you pay for. | 1¢ |
| No document found | Nothing document-shaped was in the frame. | Free |
| Unreadable | A document was there, but no usable text could be read off it. | Free |
| Unsupported document | The document was found, but its type is not one we know. | Free |
| Rejected | Recognition failed on our side — an internal error or a timeout. | Free |
Starting out
No account
10 free recognitions on the public sandbox key, with no account at all.
After you register
20 free documents when you register, credited to your balance.
Then
After that you top up the balance and every recognised document draws one credit from it.
Everything else is included at the same price: history retention you choose, the portrait crop, MRZ check digits, the cross-checks between the machine-readable zone and the visual zone, and the full field set. See the documentation for the request options and the response shape.