Analysis complete
This is an indicative review of the text sample.
A free AI text detector for a clear, quick second opinion.
Paste a sample or open a plain-text file.
This is an indicative review of the text sample.
Sentence-length variation
Variety across the sample
Use this as one clue, not a final verdict.
The result is a heuristic text check, not proof or a probability of authorship.
LexyAI is a free AI text detector and AI writing checker for essays, articles, reports, emails and everyday writing. Instead of presenting a long list of technical metrics, it combines measurable features of the sample into one readable score, supported by sentence-rhythm and vocabulary indicators. The result is a mathematical second opinion about writing patterns, not a claim to know who wrote the text.
Paste a passage, type directly into the editor or open a plain-text file. LexyAI reviews the sample and gives you a clear first look at patterns that can appear in generated, heavily edited or naturally structured writing.
The main 0–100 AI-like pattern score brings several measurable text features together. Two supporting indicators make the result easier to read: sentence rhythm and vocabulary variety. The number is an index, not a probability of authorship.
Writing style is complex. Sentence lengths, repeated phrases, word distribution and punctuation can each tell a different part of the story, so LexyAI considers multiple signals instead of relying on a single phrase or writing habit.
Use the AI writing checker for a first review of essays, articles, reports, product descriptions, emails and drafts. Compare the result with the source, the writing context and other evidence before drawing conclusions.
An AI text detector cannot prove who wrote a passage. Formal writing, editing, translation, short samples and unfamiliar language patterns can all affect the result. Never use one score as the sole basis for academic, workplace or other consequential decisions.
The detector looks at measurable properties of the text itself. It does not search for a hidden watermark or match the passage against a library of chatbot answers. Instead, it combines several statistical and linguistic signals into an indicative score that is easier to understand than a technical report.
Sentence lengths are compared across the sample. A text with a very even rhythm can look different from one that alternates between short and long sentences, but either style can be human or AI-assisted.
The analysis reviews how often words repeat and how varied the vocabulary remains across nearby sections. Topic, terminology and writing purpose all influence this measure.
Recurring word pairs and familiar transitions are treated as one signal among several. A repeated phrase is not evidence by itself: people naturally reuse terms, especially in focused or technical writing.
Local word-distribution changes and punctuation spacing add context about how the writing varies from one part to another. These statistical features are interpreted together rather than as standalone rules.
A higher score means that more of the measured features overlap with patterns the heuristic associates with generated or highly standardized writing. A lower score means fewer such patterns were measured. A middle score reflects mixed signals. None of these outcomes confirms whether a person used an AI tool.
Formal essays, formulaic reports, translated passages, edited drafts and writing by people who prefer consistent sentence structures may produce patterns that look unusual to a detector. AI text can also be revised extensively. Always review the actual writing and its context.
Use a continuous passage rather than a title, a few sentences or disconnected bullet points. LexyAI requires 50 words, while a longer sample usually provides more material to compare. Check a second passage when the first one is very short or highly specialized.
An AI writing detector and a plagiarism checker answer different questions. LexyAI does not search websites, books or academic databases for matching passages. It reviews statistical features of the submitted text and cannot establish originality or ownership.
No. LexyAI offers an indicative review of writing patterns, not proof of authorship. Human writing, edited text and AI-assisted writing can overlap.
The 0–100 score is an internal heuristic index that summarizes measured text patterns. It is not a calibrated probability that a person or AI system wrote the sample.
LexyAI reviews sentence-length variation, vocabulary variety, repeated word pairs, local word-distribution patterns, punctuation rhythm and selected familiar phrases.
At least 50 words are required. A longer, connected sample gives the analysis more context; very short samples should be interpreted especially carefully.
No. LexyAI does not identify a specific model, chatbot or author. It reviews patterns in the submitted text.
No. It does not compare writing against published sources or assess whether text has been copied. It reviews selected statistical and linguistic patterns.
The current analysis runs in a browser Web Worker. The submitted text is processed in the page and is not sent to an analysis server.
No detector should be assumed to work equally across languages and writing styles. LexyAI uses general text-pattern measures and selected English phrase signals, so results outside English should be treated cautiously.