Sign in

v1

latestOpenAPI 3.1.02026-07-2219165126.1 KB
Moderations

Create a moderation

Classify text or image inputs against moderation categories and scores.

post/v1/moderations

Request body

modelstring required

Moderation model identifier.

Example request

{
  "input": [
    {
      "text": "I want to hurt someone.",
      "type": "text"
    },
    {
      "image_url": {
        "url": "https://example.com/image.png"
      },
      "type": "image_url"
    }
  ],
  "model": "omni-moderation-latest"
}

Response

Successful Response

idstring required
modelstring required

Example response

{
  "id": "mdr_123",
  "model": "omni-moderation-latest",
  "results": [
    {
      "categories": {
        "self-harm": false,
        "violence": true
      },
      "category_applied_input_types": {
        "violence": [
          "text"
        ]
      },
      "category_scores": {
        "self-harm": 0.04,
        "violence": 0.92
      },
      "flagged": true
    }
  ]
}