Skip to main content
POST
/
classification
Classifies a Document into labels based on its content.
curl --request POST \
  --url https://api.totoy.ai/v1/classification \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "rdc-de-2025-02-14",
  "document": "JVBERi0xLjUNCiW1tbW1DQox..."
}
'
{
  "label": {
    "reasoning": "Considering the title 'MR-Zuweisung' and the contents, particularly the fact that it is a referral for MRI examination due to specific clinical symptoms and conditions noted, the most appropriate classification is 'Zuweisung'. The document fits perfectly as a referral for imaging based on diagnostic requirements set by a physician.",
    "prediction": "Zuweisung"
  },
  "review": {
    "reasoning": "The document is clearly legible and formatted as a professional referral, indicating no need for further review. The content is straightforward and matches the 'Zuweisung' classification.",
    "recommended": false
  }
}

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json
​
model
enum<string>
required

The name of the model that will be used for the Classification. A model name without a date is an alias to the latest version of that model, i.e. rdc-de always points to the latest version of rdc-de.

Available options:
rdc-de-2025-02-14,
rdc-de
Example:

"rdc-de-2025-02-14"

​
document
string<byte>
required

Base64-encoded contents of a document (PDF, JPG or PNG)

Example:

"JVBERi0xLjUNCiW1tbW1DQox..."

Response

OK

​
label
object

The label

​
review
object

Review recommendations for this field