IDPARSE
Start free trial

Document data extraction API

Read the document. Keep nothing.

Turn passports, ID cards and visas into data your application can use. Document content and extracted fields are not saved in our records or logs.

Start free trial Read the docs
Free trial: 100 scans in total. No card needed.
A document's short stayPoint at a field on the document to find it in the response.Illustration
UTOPIAPASSPORTSPECIMEN
Type
P
Code
UTO
Passport no.
L898902C3
Surname
ERIKSSON
Sex
F
Given names
ANNA MARIA
Nationality
UTO
Date of birth
12 AUG 1974
Date of expiry
15 APR 2012
P<UTOERIKSSON<<ANNA<MARIA<<<<<<<<<<<<<<<<<<< L898902C36UTO7408122F1204159ZE184226B<<<<<10
Illustrated cleanupAn illustration of the retention policy
Response sent to your app
Response excerpt. See the full response in the docs
00document characters
being processed

Animation: a specimen passport is read, the extracted fields appear in the response, then the document and its machine readable zone dissolve until nothing is left. The response stays.

The machine readable zone (MRZ), as the engine reads it from the upload

One API.
Three familiar documents.

PassportsTD3
ID cardsTD1 / TD2
VisasMRVA / MRVB

One request, one honest answer

Send one file. Get the fields back as JSON (a common text format for data). The response includes check-digit results and, where available, a failure reason. Reading errors can remain even when checks pass.

  • Photos, scans and PDF filesJPEG, PNG, HEIC (the iPhone photo format), TIFF, WebP and PDF. The file type is detected from the content, not from the file name.
  • "No MRZ found" is not chargedIf the file contains no machine readable zone, the scan does not count against your included scans. Per-minute request limits also apply to unsuccessful attempts.
  • Processing time variesProcessing time depends on the file and current server load. Extraction results include the engine's processing time.
See every failure reason in the docs
Request
curl https://api.idparse.com/v1/extract \
  -H "Authorization: Bearer idp_live_…" \
  -F "file=@passport.jpg"
Response when the photo is too poor200
{
  "success": false,
  "documents": [],
  "document_count": 0,
  "failure_reason": "mrz found but check digits failed",
  "duration_ms": 910
}

Zero retention, by architecture

IDParse has no document archive. This is not a setting that we could switch on. This is the full life of a document inside IDParse:

  1. Upload

    Your file reaches the public API over an encrypted connection. Processing can use temporary files.

  2. Read and check

    The engine looks for the machine readable zone, reads it and checks the available check digits. Nothing from the document is saved in our records or logs.

  3. Respond

    The API returns the extraction result to your application as JSON, a common text format for data.

  4. Remove

    The software is designed to remove temporary files during cleanup after processing.

Examples of request information we keep

  • The number of requests
  • The time of each request
  • The processing duration
  • The outcome code (for example "success" or "no MRZ found")
  • The file size

What we never keep

  • The document image or PDF
  • Names, numbers, dates, or any other extracted field
  • File names
  • Document data in logs or error reports
  • Documents or extracted fields for software training
Our retention rule, the same for every request and every plan.
Document content and extracted fields are not added to our application records or logs.

The privacy policy lists the account, billing and technical data we keep.

What happens to temporary files

Some processing steps use temporary files. The software is designed to remove them during cleanup. They are not added to our account records or application logs.

Where documents travel

We process documents on dedicated servers in Germany, hosted by Hetzner. Results are returned to your application.

Check digits help detect reading errors

Some fields in the machine readable zone have check digits. IDParse recalculates these digits and reports whether they match. A matching digit is a consistency check, not proof that the field is correct.

Select a field above. Check digits help detect reading errors in the document number, birth date and expiry date. The name has no check digit of its own, so IDParse reports it without a validity flag.

What IDParse reads, and what it does not do

DocumentMRZ formatReturned as
Passports2 × 44TD3
ID cards3 × 30TD1
Older ID cards and residence permits2 × 36TD2
Visas2 × 44, 2 × 36MRVA, MRVB

IDParse extracts data. That is all it does.

  • It does not decide whether a document is genuine.
  • It does not compare faces or create biometric templates.
  • It does not offer a document archive.
  • It does not claim to be right every time. It tells you which fields passed their check digit.

Plans with included scans.

Every plan stops at its included scans, so there are no extra usage charges. You decide yourself when to move to a larger plan.

Free trial €0 100 scans in total Starter €39 1,000 included scans Growth €99 5,000 included scans Business €299 20,000 included scans
Prices per month in euro. No value added tax (VAT) is added. Compare plans

Try it with a document you are allowed to process.

The free trial has 100 scans in total. No card needed.

Start free trial Read the docs