BotOrNot

Don’t gamble on one detector.

Run a quick sample across multiple detectors and see where they agree—or disagree.

Multi-strategy detection
Responsible use: AI detection is imperfect. Use multiple signals, not a single score.
Aggregate read
—
Run a scan to see a confidence band.
PDF export and rationale are Pro features.
We respect your privacy
We don’t store your content or use it for training. Delete anytime.

How our multi-detector system improves reliability

Single detection methods often disagree. We use multiple specialized AI-powered strategies analyzing different aspects (patterns, perplexity, word-level, classification, binary verdicts) and aggregate results to reduce errors.

Consensus & spread
We display an aggregate verdict with a confidence band based on cross-detector spread. Larger spread → more caution.
Short-text guardrails
Many detectors struggle below ~150 words. We warn you when reliability drops and suggest alternatives.
Evidence Pack (Pro)
Export a PDF/CSV with per-detector scores, rationale, and a plain-English interpretation for audit trails.

Academics: use results responsibly

Guidance for students, teachers, and professors using BotOrNot in learning environments.

Students
  • Run drafts to spot detector disagreement before submission.
  • Attach Evidence Pack to conversations with instructors.
  • Short texts? Add context or combine sections for better signals.
Teachers
  • Batch-check up to 5 files with Educator Pro.
  • Use class summaries to find outliers, not to auto-penalize.
  • Share the report when contacting students; prefer revision options.
Writing centers
  • Use disagreement to coach revisions (tone, pacing, citations).
  • Keep a record with PDF/CSV exports.
  • Adopt our fairness checklist in your policy.

FAQ

How accurate is this?
AI detection is imperfect. That's why we aggregate multiple specialized detection strategies and show disagreement. Use results as one signal among many, not a verdict.
Do you store my content?
No. We don't store or train on your content. You can delete any artifacts anytime from your dashboard.
What's included in Pro?
Per-detector rationale, confidence bands, PDF/CSV exports, submission history, and priority speed.
Which AI detectors does BotOrNot use?
BotOrNot uses multiple specialized detection strategies powered by advanced AI models: Pattern Analysis (linguistic naturalness), Token Perplexity (probability patterns), Word Analysis (AI word detection), Multi-Class Classification (human/AI/mixed/paraphrased), and Binary Classification (confident AI vs Human verdicts). These strategies are aggregated to provide the most reliable detection results.