Contract and General Document Reading

What it does

When a document does not match a known type, the universal reader still returns detected type, readable text and any label/value pairs it can identify — so nothing is a dead end.

How it works

1. 1Upload the document with no doc_type
2. 2Type detection runs first
3. 3The universal prompt extracts whatever structure exists
4. 4Anything unstructured is returned as text

Sample result

Illustrative response shape using synthetic data.

{
  "type": "Contract",
  "values": { "document_type": "contract" },
  "field_status": { "document_type": "read" }
}