Text Detection
Detect AI-generated or AI-assisted text with sentence-level analysis
Results will appear here
Upload content and click Analyze
Detect AI-generated or AI-assisted text with sentence-level analysis
Results will appear here
Upload content and click Analyze
TruthLens is a free AI-generated content detection platform built for everyone — not just researchers or developers. Whether you're a student, journalist, educator, or curious reader, TruthLens helps you verify whether text or images were created by an AI tool.
Unlike other tools, TruthLens provides sentence-level analysis so you can see exactly which parts of a text are likely AI-written — not just a single score for the whole document.
📝
Text Detection
Detects ChatGPT, Claude, Gemini, Llama and more
🖼️
Image Analysis
Forensic inspection with ELA and noise analysis
🔍
Sentence Level
See exactly which sentences are AI-written
🔒
Privacy First
Your content is never stored on our servers
Every number here reflects a real design decision we made to build something trustworthy.
0%+
Text detection accuracy
On modern LLMs
0
Detection signals
Per text analysis
0
Forensic methods
For image analysis
0
Content stored
Your privacy guaranteed
A combination of machine learning models and established forensic techniques working together.
Our text detector uses a fine-tuned DeBERTa-v3 model trained on millions of human and AI-written samples.
Perplexity Analysis
AI text is highly predictable — low perplexity scores indicate machine generation
Burstiness Scoring
Humans write in bursts of short and long sentences. AI writes uniformly
Sentence Classification
Each sentence is scored individually so you can pinpoint AI-written sections
DeBERTa-v3 Model
State-of-the-art transformer trained on GPT-4, Claude, Gemini and Llama output
Three research-backed forensic techniques alongside a vision transformer model to detect signs of AI generation.
EXIF Metadata Forensics
Real photos contain camera metadata. AI images often lack it or show editing software traces
Error Level Analysis (ELA)
Based on Krawetz (2007) — detects recompressed or edited regions by comparing compression artifacts
Noise Pattern Analysis
Real cameras produce stochastic sensor noise. AI images have unnaturally smooth patterns
Image Type Detection
Identifies photos, illustrations, screenshots and documents — applying the right analysis for each
From students to journalists, TruthLens is designed for anyone who needs to verify content authenticity.
🎓
Verify whether submitted assignments were AI-written. Promote academic honesty without accusing students unfairly.
📰
Quickly verify the authenticity of quotes, articles, and images before publishing or sharing breaking news.
🏢
Screen job applications, verify vendor documents, and ensure the content you receive is genuinely human-created.
🧑💻
Use our API to integrate AI detection into your own products, platforms, and research pipelines.
✍️
Prove your content is human-written and build trust with your audience in a world full of AI-generated material.
👤
Anyone who wants to know if what they're reading or seeing online was created by a human or an AI tool.