Structured data from any document
Extract typed, schema-validated fields from IDs, passports, invoices, and forms. Each field comes with a confidence score so you can automate with confidence or escalate edge cases.
- 98.7% field extraction accuracy
- Per-field confidence scores
- All ICAO passport and ID formats
- ~240ms median latency
Live demo — real Claude extraction
Upload an ID or document to see structured field extraction in real time.
Input document
Drop an ID, passport, or form
or click to browse · JPEG, PNG, WebP · max 5MB
Extracted fields
Upload a document to extract fields
Production-ready extraction
Structured JSON output
Every extraction returns schema-validated JSON. No regex, no post-processing.
Per-field confidence
Know exactly which fields are reliable. Route low-confidence cases to human review.
Supports all ID types
Passports, driver's licenses, national IDs, invoices, birth certificates, visas.
Custom schema
Define your own field schema for domain-specific forms and contracts.
Typed fields, zero parsing
Stop writing regex for document fields. Quantilence returns structured JSON with the right types, validated against a schema, with per-field confidence so your pipeline knows exactly what to trust.
curl -X POST https://api.quantilence.com/v1/document-ai/extract \
-H "Authorization: Bearer qk_live_..." \
-F "file=@passport.jpg"{
"documentType": "passport",
"overallConfidence": 0.987,
"latencyMs": 241,
"fields": {
"fullName": { "value": "Jane Smith", "confidence": 0.99 },
"dateOfBirth": { "value": "1990-04-21", "confidence": 0.98 },
"documentNumber": { "value": "A1234567", "confidence": 0.99 },
"expiryDate": { "value": "2030-04-20", "confidence": 0.99 },
"issuingCountry": { "value": "USA", "confidence": 0.97 },
"nationality": { "value": "American", "confidence": 0.96 },
"gender": { "value": "F", "confidence": 0.99 },
"mrz": { "value": "P<USASMITH<<JANE<<<...", "confidence": 0.95 }
}
}