Image Tools

Compress, resize, HEIC→JPG, crop, ID photo, QR, OCR, background remover — all processing in your browser. Files never leave your device.

Files never leave your device. Uploaded images and PDFs are processed only in your browser's memory — even the site operator cannot see their contents. Safe for resumes, IDs, and sensitive documents.
[ AdSense — Image tools category ]
📸 Image Tools
9
🗜️
Image Compress
Quality slider to shrink files — fits Gmail / messaging attachment limits
📐
Image Resize / Format Convert
Resize by px or % + JPG/PNG/WebP format conversion
📷
HEIC → JPG
Batch convert iPhone / iPad photos to JPG or PNG
🖼️
Image Crop
Social ratio presets (1:1·4:3·16:9·9:16) + free crop + rotation
🪪
ID / Passport Photo
Standard sizes (US passport, Schengen, LinkedIn) at 300 DPI auto-crop
▣
QR Code Generator
URL · text · Wi-Fi → PNG / SVG download
🔍
QR Code Scanner
Extract text / URL / Wi-Fi from camera or image
🔡
Image OCR
English & Korean OCR — Tesseract.js WASM, in-browser
✂️
Background Remover
AI auto-removes photo backgrounds → transparent PNG (small model ~43MB)

— Processed only in your browser

Photos never leave your device — safe for resumes, IDs, and sensitive images.

Cloud-based tools like TinyPNG and remove.bg upload your images to their servers; this site does everything locally in your browser memory and via WebAssembly. Image compression, resizing, HEIC→JPG conversion (iPhone photos), cropping (social aspect ratios), ID/passport photos, QR generation/scanning, OCR, and AI background removal — 9 free image tools, no signup, no install. Only the OCR (~10–13MB) and background remover (~43MB) AI models are auto-downloaded on first use, then cached. PDF merge, split, edit, and image↔PDF conversion live in the separate PDF tools category.

Last updated: 2026-05-09 9 tools No upload

Frequently Asked Questions

Are uploaded files really not sent to any server?
Correct. All processing happens with JavaScript and WebAssembly in your browser. Image compression uses the canvas API, PDF uses pdf-lib (JS), OCR uses Tesseract.js (WASM), and background removal uses ONNX Runtime Web (WASM) — all client-side. Even the site operator cannot see the contents of your files. Only the OCR training data and background-removal AI model are auto-downloaded from jsdelivr / staticimgly.com on first use.
How accurate is OCR?
Tesseract.js typically achieves 90–95% accuracy on clean printed text. Handwriting, cursive, complex tables, and vertical text drop to 50–70%. Accuracy depends heavily on image resolution (300dpi+ recommended), background contrast, and font.
Can it handle large files?
On desktop Chrome, 100MB+ images and PDFs work fine. iOS Safari and mobile browsers may become unstable above ~500MB due to memory limits. For PDF merge and other multi-file tools, total combined memory matters.
Are AI models re-downloaded every visit?
No. Models download once and are cached in IndexedDB and the browser HTTP cache, so subsequent visits start instantly. OCR English/Korean ~10–13MB, background remover small model ~43MB. Wi-Fi recommended for first run.