- Home
- AI Content Detectors
- AI-Content Detector ML
AI-Content Detector ML
Morsbach, Germany
What it is
AI-Content Detector ML is a free browser extension and web tool that estimates whether text was generated by ChatGPT, Claude, Gemini, Llama, and other AI models. It recognizes GPT-3, GPT-4, Claude, Gemini, and Llama with color-coded badges and shows live detection confidence. Text can include social posts, articles, comments, and emails. German, English, and Vietnamese are supported with automatic language detection. The machine-learning model was trained on 15,234 human and AI samples; the maker reports about 92% accuracy on test data and notes that results vary by length, language, and complexity. At least 50 words are recommended, since shorter text may produce lower confidence. After an approximately 25 MB model download, analysis runs locally in the browser. The maker says data is not sent to servers or third parties, while settings and a whitelist remain local. Dark mode and an AI-content blur effect are included.
For and against
For
- Free browser extension and web tool
- Supports three languages
- Runs analysis locally after download
- Shows confidence and model badges
- Supports common text formats
Against
- Shorter than 50 words may reduce confidence
- Reported accuracy varies by text and language
- Requires an initial model download
The verdict
AI-Content Detector ML combines local browser processing with model labels and confidence indicators. Its language and short-text limits are important when interpreting results.
Who it suits
It suits people checking text in supported languages who want browser-based AI-origin estimates. Users handling short passages should account for the stated confidence limitation.
Plans and prices
AI-Content Detector ML
completely free; analyses run locally in browser
Read 4 Oct 2026 · content-detector.ai
On the record
- Browser availability
- The download section lists Chrome Web Store, Firefox Add-on, and Edge Add-on, and names Chrome 88+, Firefox 109+, Edge 90+, Opera 75+, and Brave as system requirements.content-detector.ai · read 4 Oct 2026
- Detected models
- It identifies GPT-3, GPT-4, Claude, Gemini, Llama, and other AI models with color-coded badges.content-detector.ai · read 4 Oct 2026
- Features
- Features include dark mode, a blur effect for AI content, and a live display of detection confidence.content-detector.ai · read 4 Oct 2026
- Languages
- It supports German, English, and Vietnamese, with automatic language detection.content-detector.ai · read 4 Oct 2026
- Offline use and privacy
- After the initial model download of about 25 MB, analysis runs locally in the browser and the maker says no data is sent to its servers or third parties.content-detector.ai · read 4 Oct 2026
- Purpose
- The product detects text generated by ChatGPT, Claude, Gemini, and other AI models.content-detector.ai · read 4 Oct 2026
- Reported accuracy
- The maker reports about 92% accuracy on its test data and says results vary by text length, language, and complexity.content-detector.ai · read 4 Oct 2026
- Short-text limitation
- The maker recommends at least 50 words for reliable results and says shorter text may have lower confidence.content-detector.ai · read 4 Oct 2026
- Stored data
- The maker says only settings and the whitelist are stored, locally in the browser.content-detector.ai · read 4 Oct 2026
- Support
- Users can report false detections through the contact form or by email, and the maker says it improves the model using user feedback.content-detector.ai · read 4 Oct 2026
- Supported content
- The detector works with text-based content including social media posts, articles, comments, and emails.content-detector.ai · read 4 Oct 2026
- Technical details
- The product uses TensorFlow.js and lists 19 features, 20 training epochs, and a WebGL/CPU backend.content-detector.ai · read 4 Oct 2026
- Training
- Its machine-learning model is trained on 15,234 samples from humans and various AI models.content-detector.ai · read 4 Oct 2026
