Free AI Text Detector
An AI text detector estimates the probability that a piece of text was written by an AI like ChatGPT, Claude, or Gemini. This detector uses transparent heuristics (sentence-length variance, lexical diversity, common AI phrasing, punctuation patterns) and tells you exactly which signals contributed to its score. It is free, runs entirely in your browser, and is not 100% accurate - use the score as one signal alongside your own judgement, never as the sole basis for an accusation.
Honest disclaimer. No AI detector is 100% accurate, including this one. False positives happen with polished, formal, or non-native English writing. Treat the score as a starting point for a conversation, not as proof.
Text is processed entirely in your browser. Nothing is uploaded.
Frequently asked questions
How accurate is this AI detector?
No AI detector is 100% accurate. This one uses transparent heuristics (sentence-length variance, lexical diversity, common AI phrasing, punctuation patterns) and reports an estimate, not a verdict. OpenAI shut down its own detector in 2023 because of low accuracy. Use this score as one signal alongside your own judgement, never as the sole basis for an academic accusation.
What does the detector look for?
It analyses four signals: (1) Burstiness - human writing varies sentence length more than AI; (2) Lexical diversity - the type-token ratio of unique words to total words; (3) Pattern phrases - frequency of phrases like "in conclusion", "it is important to note", "delve into", "navigate the complexities"; (4) Punctuation density - em-dashes, semicolons, and Oxford commas appear at distinctive rates in AI output.
Why does my human-written text get flagged as AI?
Polished, formal writing - exam essays, business memos, technical documentation - shares many signals with AI output: even sentence lengths, formal vocabulary, low slang. False positives are most common with non-native English writers, edited drafts, and academic writing. Always combine the score with your own judgement.
Will my text be sent anywhere?
No. The detector runs entirely in your browser using JavaScript. The text you paste never leaves your device, never touches our servers, and is not stored anywhere. Close the tab and the text is gone - this is one of the few private AI detectors on the web.
Why is this free when most AI detectors charge?
Because the heuristics are public knowledge and the math runs in your browser. Paid detectors run cloud GPU inference on every request, which costs money. We trade some accuracy for full privacy and zero cost. If you need higher accuracy and your school requires a specific tool, use Turnitin or GPTZero in addition.
How can I tell my own writing apart from AI to a teacher?
Keep drafts. Save your version history in Google Docs (View → Version history). Show outlines, handwritten notes, and the messy first draft. Cite specific sources you used. AI-generated work usually lacks this paper trail. Combined evidence is far more persuasive than any single detector score.