About Us
We built the most accurate AI text detector because we believe you deserve to know the truth about what you're reading — no black boxes, no guesswork.
The Origin Story
Our founder holds a degree in Artificial Intelligence from Belarusian State University of Informatics and Radioelectronics (BSUIR). Their diploma thesis, completed in 2024, focused on“Linguistic Methods for Detecting Machine-Generated Text.” During this research, they identified critical flaws in existing detection tools — they were unreliable, lacked transparency, and often produced false positives. The goal was to build a detector that was both accurate and explainable.
The thesis work became the foundation of our system. After graduation, the founder partnered with two fellow BSUIR alumni — a linguist and a full-stack engineer — to turn the research prototype into a production‑ready service that anyone could use for free.
How It Works
Our detection engine is built on peer-reviewed research from Stanford, UCLA, University College London, and other leading institutions. We analyze text using three core linguistic metrics:
- Perplexity: Measures how predictable the text is. AI-generated text typically has lower perplexity (more predictable) than human writing. Human median perplexity is around 35.9 vs 21.2 for AI.
- Burstiness: Analyzes sentence length variability. Human writing has natural “bursts and pauses” — a rhythm that AI struggles to replicate.
- Entropy: Evaluates word choice randomness. AI tends to use more predictable vocabulary, while humans choose words with greater variety.
We combine these metrics with an ensemble learning model (XGBoost, LightGBM, and CatBoost) that achieves 99.5% accuracy in detecting text from 10+ AI models, including GPT-5.5+, Claude 4.7+, Gemini, LLaMA, DeepSeek, and Mistral.
What We Believe
AI is transforming the world — but that transformation should happen with transparency, not secrecy. We believe:
- Truth matters. You have the right to know whether what you're reading was written by a human or a machine.
- Detection should be explainable. A score without context is useless. That's why we provide sentence-by-sentence analysis with clear reasoning.
- Privacy is non-negotiable. Your text is transmitted encrypted and never stored permanently.
- Education comes first. Teachers and students shouldn't have to pay hundreds of dollars to protect academic integrity.
Our Team
We are a small, passionate team of researchers, linguists, and engineers based in Minsk. Our backgrounds span computational linguistics, machine learning, and secure systems — all united by a common goal: making AI detection accessible, reliable, and transparent for everyone.
Trusted Worldwide
Today, our service is used by over 50,000 educators,200,000 students, 100,000 writers, and 5,000+ companies across the globe. From high school teachers verifying assignment authenticity to freelance journalists double-checking their work before publication — our tool helps people make informed decisions about the text they read and write.
We're proud to be one of the few detectors with a false positive rate under 2%, independently verified by third-party evaluations. That means fewer false accusations, fewer unfair grades, and more trust in the tools you use.
Join Us
Whether you're a teacher protecting academic integrity, a student verifying your own work, a writer ensuring authenticity, or a business screening content — we're here to help.
Try the detector today. It's free, it's fast, and it tells you the truth.