Document type
General Document OCR
Any document that does not match a known type still returns type detection, text and fields.
Can Mahad OCR read a General Document?
Available now. Any document that does not match a known type still returns type detection, text and fields. It returns document_type (detected); fields[] label/value pairs; readable text and more as JSON. Input: PNG, JPG, PDF up to 10 MB (5 pages). Limitation: Structure depends entirely on the document.
Fields normally extracted
- document_type (detected)
- fields[] label/value pairs
- readable text
- field_review (✓ / ⚠ per field)
Validation available
Type detection, ✓ / ⚠ field checks with evidence, human review and duplicate detection all apply.
Example response
Synthetic example — not a real document.
{ "type": "Certificate", "values": { "document_type": "certificate" }, "status": "verified" }Supported formats
PNG, JPG, PDF up to 10 MB (5 pages)
Common business uses
- Mixed document intake
- Certificates and letters
- Anything without a fixed template
Limitations
Structure depends entirely on the document. Expect text and partial fields rather than a guaranteed schema.
Related features
Try it on your own document
Upload a general document in the live demo — the file is processed in memory and not stored.