Documentation
Learn how to integrate and use Free AI Detect.
Welcome to the Free AI Detect documentation. This guide will help you understand our detection methodology and how to interpret results.
How Detection Works
Our system analyzes text for patterns common in AI-generated content, focusing on two main metrics: Perplexity and Burstiness.
- Perplexity: Measures the randomness of the text. High perplexity indicates human-like writing.
- Burstiness: Measures the variation in sentence structure and length. AI often has low burstiness (uniformity).
Interpreting Results
Detection scores are presented as a percentage. A higher percentage indicates a higher likelihood of AI generation. Scores below 20% are generally considered human-written.
Supported Models
We specialized in detecting content from GPT-3.5, GPT-4, Claude 3.5, Gemini 1.5, and LLaMA 3. We update our models weekly to stay ahead of new LLM releases.