Trustfull Document checks whether an uploaded payslip, tax document, bank statement or utility bill was edited, reads its content, and returns a 0–1000 risk score.

Request

The Document endpoint works asynchronously and accepts a multipart/form-data request. When you submit a request, the API immediately responds with a 200 OK status if the upload is valid and accepted for processing. The actual analysis happens in the background.

Each request_id accepts a single document submission — repeated submissions with the same request_id are rejected.

You can also send model_id to choose the model that computes the score. Use one of your custom models, or a Trustfull standard model such as document_score. Without it, the API uses the active model.

Below is an example of a Trustfull Document request:

curl --request POST \
     --url https://api.fido.id/1.0/document \
     --header 'Accept: application/json' \
     --header 'x-api-key: [your_app_key]' \
     --form 'request_id=your_request_id' \
     --form 'document_type=payslip' \
     --form 'file=@/path/to/document.pdf' \
     --form 'model_id=your_model_id'

The accepted form fields are request_id, document_type, file, and model_id. Any other field — including a customer-supplied file_type — is rejected with 400 Invalid request.

document_type is the semantic category of the upload and must be one of: payslip, tax_document, bank_statement, utility_bill.

The file content type is detected by the server from the uploaded bytes; the accepted MIME types are application/pdf, image/jpeg, image/png, image/webp (reported back as PDF, JPEG, PNG, WEBP). Anything else (e.g. .zip, plain text) is rejected. The maximum upload size is 4 MiB.

Response

A successful request returns the polling URL:

{
  "status": "ok",
  "url": "https://api.fido.id/1.0/document/your_request_id"
}

Poll that URL (see Get Document API) to retrieve the analysis result.