AI Content Detector

Overview

This Apify actor detects AI-generated content in text. It analyzes provided text or files and returns a score indicating the likelihood the content was written by AI versus a human.

Key Features

• Content Analysis: Scans text or uploaded files for AI-generated content.
• File Support: Accepts .txt, .md, .html, .pdf, and .docx files.
• Large Text Handling: Automatically splits texts longer than 50,000 words into manageable chunks for analysis.
• Detailed Scoring: Provides an overall AI/Human score and a per-text-block breakdown.
• Reliability Features: Includes proxy support and a retry mechanism.
• Structured Output: Stores results in an Apify dataset, exportable as JSON, CSV, or Excel.

How to Use

Run the actor on the Apify platform. Provide input via the textContent field for direct text, or use fileUpload for a file. If a file is uploaded, the textContent field is ignored.

Input

The actor accepts a JSON input object with the following parameters:

• textContent (string, optional): The text string to analyze.
• fileUpload (file, optional): A file to upload for analysis. Supported formats: .txt, .md, .html, .pdf, .docx.
• proxyConfiguration (object, optional): Recommended for faster, more reliable runs. Use { "useApifyProxy": true }.

Example Input:

{
  "textContent": "I was just sitting here, watching the rain tap against the window... In contrast, our synergistic platform leverages advanced natural language processing...",
  "proxyConfiguration": {
    "useApifyProxy": true
  }
}

Output

Results are saved to the Apify dataset. The output is a JSON object containing overall scores and a detailed breakdown by text block.

Output Fields:
* original: Overall human-written score for the entire input.
* ai: Overall AI-generated score for the entire input.
* blocks: An array of analyzed text segments.
* text: The segment of analyzed text.
* result: The scores for this specific block.
* real: Human-written score.
* fake: AI-generated score.
* status: Analysis status (e.g., "success").

Example Output:

{
  "original": 0.0365,
  "ai": 0.9634,
  "blocks": [
    {
      "text": "I was just sitting here, watching the rain tap against the window. ",
      "result": {
        "fake": 0.8077,
        "real": 0.1923,
        "status": "success"
      }
    }
    // ... more blocks
  ]
}