✨ AI-Powered · Runs On-Device
AI Tools
AI Face Auto-Crop
Is your face off-centre, too small, or too close to the edge of the frame? This tool uses on-device AI to detect your face, re-centre it with the correct padding for passport-style photos, and then resizes it to the exact spec for your form — Aadhar, PAN, Passport, MPSC, SSC and more.
How it works
1
Upload your photo
JPG or PNG, any size
2
AI detects & centres face
Automatic — toggle on by default
3
Pick a preset & download
Correctly sized, named & compressed
Select document type
💡 The "Auto-centre face with AI" toggle below the upload box is on by default. If your photo already has the face well-centred, you can switch it off to use the original framing instead.
Click to upload or drag & drop
JPG, PNG supported · Max 10 MB
🔒 Photo never leaves your device
Quality / Compression
80%
KB
(auto-adjusts quality to hit this limit)
Result
Ready to download
Face-cropped —
Resized —
Original
—
New size
—
Saved
—
Dimensions
—
AI Face Auto-Crop — Frequently Asked Questions
The tool uses a small on-device face detection model (TinyFaceDetector via face-api.js) that runs entirely in your browser. It finds the largest face in the photo and calculates a crop region with the correct padding above the head and below the chin — the standard framing required for passport-style photos.
If the AI can't confidently detect a face — for example in a very dark, blurry, or heavily cropped photo — it falls back to using your original photo unchanged, and you'll see a small notice under the upload box. You can still resize and download normally.
Yes. Untick "Auto-centre face with AI" below the upload box before or after uploading. The tool will re-process your original photo without face detection, using the standard centre-crop instead.
Yes. Auto-crop runs first, then the cropped photo is resized and compressed to whichever preset you choose — Aadhar, PAN, Passport, MPSC, SSC, UPSC, IBPS and more all use the same crop, just at different output dimensions and KB limits.
No. Both face detection and resizing happen locally using your browser's Canvas API and an on-device AI model. The model files are fetched once from a CDN, but your photo itself is never sent anywhere.
The first time you use this tool in a session, your browser downloads a small (~190 KB) face detection model. After that, detection on subsequent photos is near-instant.