Omni by Moderate - Content Moderation API

Overview

The Omni API returns responses in a consistent JSON format. This page explains the structure and fields in the response to help you interpret and use the moderation results effectively.

Response Structure

Here's an example of a complete response from the API:

{
  "id": "modr-123456",
  "model": "text-moderation-latest",
  "results": [
    {
      "flagged": false,
      "categories": {
        "hate": false,
        "harassment": false,
        "self-harm": false,
        "sexual": false,
        "violence": false
      },
      "category_scores": {
        "hate": 0.0,
        "harassment": 0.0,
        "self-harm": 0.0,
        "sexual": 0.0,
        "violence": 0.0
      }
    },
    {
      "flagged": false,
      "categories": {
        "hate": false,
        "harassment": false,
        "self-harm": false,
        "sexual": false,
        "violence": false
      },
      "category_scores": {
        "hate": 0.0,
        "harassment": 0.0,
        "self-harm": 0.0,
        "sexual": 0.0,
        "violence": 0.0
      }
    }
  ],
  "combined_text_moderation": {
    "flagged": false,
    "categories": {
      "hate": false,
      "harassment": false,
      "self-harm": false,
      "sexual": false,
      "violence": false
    },
    "category_scores": {
      "hate": 0.0,
      "harassment": 0.0,
      "self-harm": 0.0,
      "sexual": 0.0,
      "violence": 0.0
    }
  },
  "vision_analysis": [
    {
      "image_index": 1,
      "analysis": "This image shows a plate of beautifully presented food in what appears to be a restaurant setting. The dish looks like a well-plated steak with garnishes and side dishes. The lighting is good, the food appears appetizing, and the image is of high quality. This is perfectly appropriate for a restaurant listing or review as it showcases the food offerings. VERDICT: APPROPRIATE",
      "flagged": false
    }
  ],
  "flagged": false,
  "signature": "abcdef1234567890"
}

Response Fields

Field	Type	Description
id	string	Unique identifier for the moderation request
model	string	The moderation model used for the request
results	array	Array of moderation results for each input item
combined_text_moderation	object	Moderation result for all text items combined (only relevant when multiple text items are present)
vision_analysis	array	Array of vision analysis results for each image (only present if analyze_images is true)
flagged	boolean	Whether any content was flagged as inappropriate
signature	string	Cryptographic signature for verifying the response

Result Object

Each item in the results array has the following structure:

Field	Type	Description
flagged	boolean	Whether the content was flagged as inappropriate
categories	object	Object containing boolean flags for each category
category_scores	object	Object containing scores (0.0 to 1.0) for each category

Category	Description
hate	Content that expresses, incites, or promotes hate based on race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste
harassment	Content that expresses, incites, or promotes harassing language towards any target
self-harm	Content that promotes, encourages, or depicts acts of self-harm, such as suicide, cutting, and eating disorders
sexual	Content meant to arouse sexual excitement, such as the description of sexual activity, or that promotes sexual services (excluding sex education and wellness)
violence	Content that promotes or glorifies violence or celebrates the suffering or humiliation of others

Vision Analysis Object

Each item in the vision_analysis array has the following structure:

Field	Type	Description
image_index	number	The index of the image in the input array
analysis	string	Detailed analysis of the image content
flagged	boolean	Whether the image was flagged as inappropriate

Interpreting Results

When processing the API response, you should check the top-level flagged field first. If it'strue, then at least one item in the request was flagged as inappropriate.

For more detailed analysis, you can check the individual results in the results array. Each result corresponds to an item in the input array, in the same order.

If you submitted multiple text items, you should also check the combined_text_moderation field to see if the combined text was flagged, even if individual text items were not.

If you requested vision analysis for images, check the vision_analysis array for detailed analysis of each image.

Verifying the Response

To ensure the response came from the Omni API and hasn't been tampered with, you should verify thesignature field. See the Verify Signature page for details.

Response Format