Image Moderation

BETA

Analyze images for potential moderation issues.

POSThttps://api.safecomms.dev/moderation/imageBETA

The image moderation endpoint allows you to scan images for unsafe content such as nudity, violence, gore, and hate symbols.

Request Body

Field	Type	Required	Description
image	string	Yes	The image to moderate. Can be a public URL or a Base64 encoded string.
language	string	No	The language to use for the explanation (default: "English").
moderationProfileId	string	No	The ID of a custom moderation profile to apply.
enableOcr	boolean	No	Extract visible English text from the image (OCR). Default: false.
enhancedOcr	boolean	No	Allocates extra resources for maximum accuracy. Slower and rarely needed for standard images. Requires `enableOcr` to be true. Default: false.
extractMetadata	boolean	No	Extract EXIF metadata from the image. Default: false.

URL Restrictions

When providing a URL for the image field, please note the following restrictions:

Download Timeout: The image must download within 5 seconds.
Redirect Limit: A maximum of 3 redirects are followed.
Protocols: Only HTTP and HTTPS are supported.
Security: URLs resolving to internal or private IP addresses are immediately skipped since we cannot access them.

We enforce these restrictions to ensure a high level of quality and reliability for our service. If you have any questions or concerns, please contact our helpdesk.

Token Usage & Costs

Image moderation uses a base cost of 20 tokens per request.

Feature	Cost	Tier	Condition
Base Request	20 Tokens	Starter+	Always applied.
Non-English Language	+1	Pro+	Applied when `language` is not 'English'.
Optical Character Recognition (OCR)	+3	Pro+	Applied when `enableOcr` is true.
Enhanced OCR	+8	Pro+	Applied when `enhancedOcr` is true. Replaces standard OCR cost. Note: Slower and rarely needed for standard images.
EXIF (Metadata)	+1	Starter+	Applied when `extractMetadata` is true.

Example Request

{
  "image": "https://example.com/image.jpg",
  "language": "English",
  "enableOcr": true,
  "extractMetadata": true
}

Example Response

{
  "isClean": false,
  "confidence": 95,
  "severity": "High",
  "categoryScores": {
    "Profanity": "None",
    "Slurs": "None",
    "Sexual": "High",
    "Hate": "None",
    "Spam": "None",
    "Bullying": "None",
    "Violence": "None"
  },
  "isBypassAttempt": false,
  "issues": [
    {
      "term": "Nudity",
      "context": "Explicit content detected"
    }
  ],
  "reason": "Contains explicit sexual content.",
  "detectedText": [
    "WARNING",
    "ADULT CONTENT"
  ],
  "metadata": {
    "Make": "Canon",
    "Model": "Canon EOS 5D Mark IV",
    "DateTime": "2023:10:27 14:30:00"
  }
}