# ID Photo Maker

> Selfie in, passport-spec photo out. A portrait-matting model (downloaded once at 6 MB, then cached) removes the background on your device; the crop locks to one of 17 country presets — US, UK, EU, India, Japan, and more — that bake in head-height percentage, exact pixel dimensions, and the JPEG byte cap each consulate actually enforces. Image stays on your device throughout; segmentation is local, no cloud round-trip.


Live tool: https://lofttools.com/tools/image-tools/id-photo-maker

Category: Image Tools


## What it does

- 17 country presets — US Passport, US Driver's License, Canada, Brazil, Mexico, UK, EU/Schengen, Germany, India Passport, India Aadhaar, China, Japan, South Korea, Australia, Student ID, Employee Badge, LinkedIn / square
- Each preset bakes in head-height percentage, exact pixel dimensions, and JPEG byte cap to spec
- Portrait-matting model (MODNet INT8 quantized, 6 MB, downloaded once and cached)
- Auto background removal runs on-device (WebGPU when available, WebGL fallback)
- Background colour picker — white (default), off-white, light grey, light blue
- Pinch-zoom and drag to position the head inside the safe-zone overlay
- Front-camera capture or photo upload from gallery
- JPEG quality auto-tuned per preset's KB cap (e.g. 240 KB US, 300 KB India)
- HEIC / HEIF input supported (iPhone photos work directly)
- No facial retouching — only background segmentation (US State Department compliance angle)
- Image processing entirely local; the only network call is the one-time model download

## How it works

1. **Selfie or upload** — Front camera, or pick a photo from your device.
2. **Auto background removal** — 6 MB MODNet model runs locally — no upload.
3. **Pick document** — 17 country presets lock the crop to spec.
4. **Adjust crop & background** — Pinch-zoom to position, pick a bg color.
5. **Download** — Sized to spec, JPEG auto-tuned to KB cap.

## FAQ

### Does my photo leave my device?

No. Background removal runs in your browser. Your selfie is never uploaded to a server. The only network request is the one-time download of the segmentation model.

### Will the US State Department accept photos from this tool?

Since January 2026 the US State Department rejects photos with AI-edited facial features. This tool only segments the background — it does not retouch your face, skin, or eyes. For a US passport you should still verify acceptance against the official guidelines, since automated AI background changes are a grey area. The dimensions, file size, and head-height percentage we output match the official spec.

### Which documents are supported?

17 presets — US Passport, US Driver's License, Canada, Brazil, Mexico, UK, EU/Schengen, Germany, India Passport, India Aadhaar, China, Japan, South Korea, Australia, Student ID, Employee Badge, and LinkedIn / square profile.

### Why is the first run a bit slower?

The portrait-matting model (~6 MB, MODNet INT8 quantized) downloads on first use and is cached by your browser. On a 4G connection this takes 1-2 seconds. After that, subsequent photos process in under 3 seconds on most phones, faster with WebGPU.

### My phone camera does not open. What do I do?

Camera access requires HTTPS and explicit browser permission. If you blocked it earlier, look for the camera icon in your address bar and re-enable. As a fallback, tap "or select a photo" to use any photo from your gallery — the rest of the flow is the same.

### What backgrounds are available?

White (default for most passports), off-white, light gray (UK and several EU), and light blue (some employee badges). The original background is replaced with the solid color you pick after the AI cutout.

## Privacy — what we do not do

This tool runs entirely in the browser via WebAssembly. Your file never reaches a Loft Tools server. Specifically:

- **No upload.** The file bytes load into the browser tab's memory and process on your own CPU. Open DevTools → Network and observe zero outbound requests carrying file data while ID Photo Maker runs.
- **No AI training on your file.** Loft does not train models. We could not train on a file we cannot see.
- **No content scanning.** No virus, copyright, or content-moderation pass against your file. The bytes are not accessible to us.
- **No server-side log of file contents, filenames, or EXIF metadata.** Cloudflare edge captures URL and truncated IP for abuse defense (standard CDN behaviour). Cloudflare Web Analytics records anonymous page hits, no cookies, no PII. Nothing about your file content reaches any log.
- **No retention.** Close the tab and the file leaves browser memory. No backups exist on our side because no copy ever existed on our side.
- **No account.** No email, no signup, no auth, no telemetry tied to you.
- **Offline-capable after first visit** (PWA). Once you've loaded a tool, it caches; later sessions work without internet. For high-sensitivity files, run the tool once online to warm the cache, then disconnect before processing.

Compare with upload-based services: each transmits your file to a processing server. Even over HTTPS, each has logs, retention windows, and subpoena exposure. Loft has none of these because the server architecture does not include your file.

## More

- All tools: https://lofttools.com/tools
- Category: https://lofttools.com/tools/image-tools
- LLM index: https://lofttools.com/llms.txt
