AI Image Detection Response

The Copyleaks AI Image Detection API returns a detailed response containing the analysis summary, image information, and a detection mask, run-length encoded (RLE), that identifies AI-generated regions.

Response Properties

string

The version of the AI detection model used for the analysis.

object

Contains the detection mask, run-length encoded (RLE). This can be used to visualize the AI-detected regions of the image. On the synchronous endpoint this object is omitted entirely when the request set maskType: "none".

Show properties

array<integer>

An array of starting positions for each AI-detected segment in the flattened 1D image array.

array<integer>

An array of lengths for each AI-detected segment, corresponding to the starts array.

array<integer>

Present only when the request included maskType: "heatmap". One confidence band per segment, aligned with starts and lengths: an integer indicating which confidence threshold that region reached. Higher bands mean higher confidence. Use it to render a gradient overlay instead of a binary mask. Omitted for the default binary mask.

object

The AI score: the share of the image’s pixels classified as human-created vs. AI-generated.

Show properties

float

The share of the image’s pixels determined to be human-created. The value ranges from 0.0 to 1.0.

float

The share of the image’s pixels determined to be AI-generated. The value ranges from 0.0 to 1.0.

boolean

Indicates whether the image was flagged as AI-generated.

object

An object containing metadata about the analyzed image.

Show properties

object

The dimensions of the image.

Show properties

integer

The height of the image in pixels.

integer

The width of the image in pixels.

object

Optional metadata extracted from the image file.

Show properties

string

The timestamp (if available in EXIF data) indicating when the image was created.

string

The AI service or tool that created the image, if this information is present in the metadata.

string

The application or device used to create the image, if available in the metadata.

string

A summary of how the image was generated., if available in the metadata.

array<object>

Present only when the request included objectDetection. One entry per object located in the image. Objects that were requested but not found are absent, so the array is not aligned by position with the request - look entries up by object.

Show properties

string

The object description from the request that this entry corresponds to.

object

The Run-Length Encoded (RLE) segmentation mask for this object, in the coordinate space of imageInfo.shape.

Show properties

array<integer>

An array of starting positions for each segment in the flattened 1D image array.

array<integer>

An array of lengths for each segment, corresponding to the starts array.

object

Metadata about the scan operation itself.

Show properties

string

The unique identifier for this scan, provided by you in the request.

integer

The actual number of credits consumed by the scan.

integer

The expected number of credits for the scan.

string

The ISO 8601 timestamp indicating when the scan was created.

Example Response

{
  "model": "ai-image-1-ultra",
  "result": {
    "starts": [0, 393216],
    "lengths": [196608, 393216]
  },
  "summary": {
    "human": 0.25,
    "ai": 0.75
  },
  "isAiDetected": true,
  "imageInfo": {
    "shape": {
      "height": 1024,
      "width": 768
    },
    "metadata": {
      "issuedTime": "2025-07-23T12:44:05",
      "issuedBy": "OpenAI",
      "appOrDeviceUsed": "OpenAI-API",
      "contentSummary": "Created using generative AI"
    }
  },
  "detectedObjects": [
    {
      "object": "face",
      "mask": {
        "starts": [1024, 3072],
        "lengths": [512, 256]
      }
    }
  ],
  "scannedDocument": {
    "scanId": "my-scan-id-1",
    "actualCredits": 1,
    "expectedCredits": 1,
    "creationTime": "2023-01-10T10:07:58.9459512Z"
  }
}

Heatmap Result

When the request includes maskType: "heatmap", the result object carries one confidence band per segment in labels:

{
  "result": {
    "starts": [0, 393216],
    "lengths": [196608, 393216],
    "labels": [2, 3]
  }
}

Reference

Actions

Data Types

Response Properties

Example Response

Heatmap Result

Next Steps

AI Image Detection Guide

AI Image Detection API Reference

​Response Properties

​Example Response

​Heatmap Result

​Next Steps

AI Image Detection Guide

AI Image Detection API Reference

Response Properties

Example Response

Heatmap Result

Next Steps