UnTXT detects AI-generated passages in text. Users paste written content (essays, reports, documents) and receive a colour-coded analysis highlighting specific passages by AI writing signal type. Eleven independent detection modules are used: SPIKE, SYNTAX, FORMAL, HUMAN, TEMPLATE, FRONTED, FORENSIC, RHETORIC, PUFF, FORMULA, and LEXICAL. Each module targets a different characteristic of AI-generated writing. An optional twelfth layer, Academic Weapon (M13), uses three frontier AI models - Claude Opus 4.7, GPT-5.5, and Gemini 3.1 Pro - in a structured adversarial debate for the highest-confidence authorship analysis available.
Target Users
Role
Student
Use
Editing and improving AI-flagged writing before academic submission
Role
Teacher / educator
Use
Checking and exporting student submissions
Role
Non-academic professional
Use
Reviewing reports, applications, and documents for AI authenticity
Confirmed Public Features
Eleven-module AI detection engine with colour-coded, labelled highlights (SPIKE, SYNTAX, FORMAL, HUMAN, TEMPLATE, FRONTED, FORENSIC, RHETORIC, PUFF, FORMULA, LEXICAL)
Academic Mode - full 11-module breakdown; default for Student role
Standard Mode - HIGH RISK / MEDIUM RISK / POSITIVE; default for Teacher and Professional roles
Analysis history with date, word count, risk score, and credit cost per entry; all stored text encrypted at rest with AES-256-GCM
PDF analysis report export (Teacher and Professional roles, A4 formatted)
CVT credit gifting - transfer unused credits to any registered user (one-hop only)
Role-based interface adapting defaults and features per selected role
Sentence-Scope Editor - full-sentence replacement suggestions shown in context for any flagged passage; accept, modify, or dismiss
Invite & Earn referral program - personal referral link for every user; both referrer and invitee receive bonus CVT credits on first analysis
Confidence Guard - stricter role-aware detection thresholds for Teacher and Professional roles, reducing false positives in review workflows
Deferred Suggestions (Student Standard Mode) - students see a "Find Replacements" button for each flag rather than auto-fetched suggestions; encourages self-correction and deeper understanding before accepting AI-writing fixes
Academic Weapon (M13) - adversarial frontier analysis using Claude Opus 4.7, GPT-5.5, and Gemini 3.1 Pro; structured multi-round debate with per-model reasoning; AUTHENTIC (human-voice) and SUSPECT (AI-flagged) consensus highlights; 4-step first-time tutorial; CVT deducted automatically based on actual token usage
Multi-language interface - English (default), Arabic (full RTL layout), Chinese Simplified, and Korean; language auto-detected from browser; manual override via globe button in nav; preference persisted per user
Pricing Model
Free tier
First 5,000 words on sign-up - no card required
Credit unit
CVT (Contextual Value Token)
1 CVT covers
Approximately 75 words of analysis
Purchase model
One-time credit bundles via Stripe - not a subscription