Detects faces in one uploaded photo with the on-device BlazeFace model (TensorFlow.js, downloaded lazily on first run) and blurs every detected box at a fixed 15px. Exports PNG as blurred_faces.png with a face-count toast — free to start with no signup.
Runs on TensorFlow.js
Content reviewed October 2026 · About Toolzum
Every face covered automatically: upload a group photo and BlazeFace finds each face on your device, pixelated into 12px+ blocks that cannot leak back. Pixelate, soft-blur, or solid-cover modes, PNG export, refine-in-editor link. Free, no signup, nothing uploads.
Detect and blur or pixelate faces across whole photo batches with on-device BlazeFace (one shared model download for the batch). Faces need confidence above 0.5 and size above 8px; faceless files are skipped with an error. Outputs PNG as <name>-anonymized.png — local processing, 2 free batches a day.
Add one custom text caption onto a photo with color, size (10–300px slider), and X/Y position controls, composed on fixed social-size canvases — Instagram Post 1080×1080, Reel 1080×1920, YouTube Banner 2560×1440, Twitter Header 1500×500, or Custom. Exports PNG locally in your browser on Canvas; 3 free downloads a day.
Composite a face from one photo onto another with manual position, scale, and blend controls on canvas. Meme makers and designers align swaps by hand — runs locally, free to start with unlimited saves.
Renders pasted HTML and CSS markup (no file upload, no URL screenshots) into downloadable PNG, JPEG (fixed quality 0.95), or SVG via the html-to-image library. Preview is DOMPurify-sanitized — free to start with no signup.
The BlazeFace model (via @tensorflow-models/blazeface plus @tensorflow/tfjs, lazy-loaded on first run) predicts face boxes on a canvas copy; each box is redrawn through a fixed ctx.filter blur(15px). Model weights download from a CDN on first use — your pixels never upload.
No — intensity is fixed at 15px with no slider. Every detected face gets the same blur in one pass.
Every box the model predicts is blurred in the same pass and the toast reports the count (e.g. Blurred 3 faces!). Zero predictions shows a No faces detected error and downloads nothing.
Any single browser-decodable image. Output is always PNG, saved as blurred_faces.png.
Face detection and blurring happen locally in your browser with the on-device model. No image data is transmitted to external servers.
Bulk Face Anonymizer
Detect and blur or pixelate faces across whole photo batches with on-device BlazeFace (one shared model download for the batch). Faces need confidence above 0.5 and size above 8px; faceless files are skipped with an error. Outputs PNG as <name>-anonymized.png — local processing, 2 free batches a day.
Bulk EXIF Stripper & Injector
Strip all EXIF metadata from JPEG, PNG, and WebP photos by re-encoding each to JPEG at quality 0.95 (files renamed *-noexif.jpg). Strip-only — no inject mode exists. Agents clean 500 listing shots before upload — local canvas work, 2 free batches a day.