Image detection
Click “Scan visible images” and get on-image labels with confidence and source tags.
Chrome Extension • Privacy-first • BYO key optional
Detect AI-generated images and text in one click.
Scan visible images or selected text directly in your browser. Local mode keeps data on-device. Boost mode lets users bring their own Gemini, OpenAI, or Hugging Face key for stronger accuracy.
- Click-to-scan UX (no auto-scan noise)
- Local ONNX inference planned for free tier
- BYO provider keys for power users
What the product is
AI Content Detector helps people quickly validate online content without changing how they browse.
Text detection
Highlight text and run “Detect AI in selection” for fast pass/fail + uncertainty handling.
Practical trust signals
Avoid fake certainty with confidence thresholds and an explicit “Uncertain” state.
How it works
Local mode (default)
Runs in-browser inference using ONNX via transformers.js in an MV3 offscreen document. No maintainer backend and no telemetry by default.
Boost mode (optional)
Users who want higher accuracy can paste their own API key. Calls go directly to the selected provider.
Current status
Phase 0: Cleanup and architecture reset complete.
Phase 1: TS/React extension shell complete.
Phase 2 next: Real local image model integration.
Architecture & Documentation
Need technical depth? The docs cover product architecture, roadmap, and implementation status.