The pipeline,expanded.
Four tools. One mission. Every humanizer runs entirely in the browser — no uploads, no servers, no leaks. Here's what's being built and when.
Image Humanizer
v3Next-generation photo pipeline
A full rebuild of the humanization engine — more layers, deeper frequency analysis, and a model-signature reader that auto-calibrates treatment strength before the first pixel changes.
- 14-layer pipeline (up from 10)
- Auto model-signature detection
- Adaptive FFT disruption v2
- 200+ camera EXIF profiles
Video Humanizer
v1Full-length video processing
Graduating from early access (v0.1) to a stable release. Frame-accurate sensor noise, temporal PRNU coherence across the full clip, and platform-specific codec presets out of the box.
- Full-length clip support
- Temporal coherence across frames
- Per-platform codec presets
- Cancelable export flow
Text Humanizer
v1Defeat AI writing detectors
Graduating from the live early-access build (v0.1) to a stable release. Expanded linguistic fingerprint rewrite, more rhythm/burstiness control, and additional detector-targeted modes — still nothing ever leaves your device.
- Zero-upload text processing
- Linguistic fingerprint rewrite
- Detector-targeted modes
- Tone & style preservation
Music Humanizer
NEWAudio anti-detection pipeline
Browser-side audio processing that strips AI generation artifacts from music and vocals. Micro-timing variation, spectral fingerprint cleaning, and realistic room acoustics injection.
- Spectral artifact removal
- Micro-timing humanization
- Room acoustics injection
- Batch WAV / MP3 export
Tools ship continuously.Start with v3 today.
Every release lands automatically in your browser. No downloads, no installs — just open the hub and run.