True redaction removes the text and images from the file — not a black box on top. Plus real in-place text editing, fill + sign, reorder pages, and OCR.
LaunchEdit text in place, sign, redact, fill forms, and reorganize PDF files in your browser — all on your device. No upload, no subscription.
Drop two versions of a contract or drawing set. Moved notes read as moves, changed values light up digit by digit, and hidden text is named.
LaunchSide-by-side or overlay PDF diff driven by one engine that reads each file into a structured model — text, vector geometry, images, form fields, annotations — and compares the models. Moves show as moves, edits show only where they happened, and any PDF type works without picking a mode.
Drop a Gerber ZIP, an ODB++ .tgz or an IPC-2581 .xml and inspect every layer — compare revisions, snap-to-pad measure, run DRC and preview the board in 3D.
LaunchOpen Gerber, Excellon, ODB++ and IPC-2581 PCB files. Inspect layers, measure traces, run DRC, export to PNG / SVG / DXF.
Open a .step / .stp part and it measures itself: volume, thinnest wall, draft and undercuts on load, and a hole list with a BOM in one click. Measure, section, export STL.
LaunchOpen a STEP / STP part and it measures itself — volume, thinnest wall, draft and undercut checks on load, plus hole detection with a BOM. Measure, section, export STL. Rust + WASM B-rep kernel, runs in your browser.
Compress, merge, split, convert, and manipulate PDF files.
Resize, compress, convert, and transform images.
Convert, compress, and cut audio and video — MP3, WAV, MP4, WebM, Opus — all in your browser, nothing uploads.
Simulate a canning process down to its F0 sterilizing value.
Encrypt files, share encrypted notes as links, and read passport IDs and Windows event logs.
Analyze packet captures, move files between devices, and console into network gear — nothing uploads.
Scan receipts into an expense ledger — your receipts and spending stay on your device.
View, measure, and repair CAD and 3D files — STEP, STL, OBJ, PLY, GLB, 3MF, DXF, and Gerber — right in your browser. Nothing uploads.
Skip the photo studio and skip the App Store. A browser-based ID photo maker turns a selfie into a passport-spec photo on any phone or laptop. Here is what works, what the per-country rules look like, and where this approach still falls short.
The first time most people need a passport photo is the worst time to figure out how to get one. The renewal letter shows up two weeks before your trip. The pharmacy print kiosk wants $15 and a forty minute window. The free app you found wants an account, a subscription, and a watermark removed in-app for $4.99.
You can get a usable passport photo at your document’s exact size without any of that — using nothing but the browser already on your phone. This post walks through how, what the per-country rules actually require, and the few cases where the studio is still the right call.
Open the ID Photo Maker on your phone or laptop. Pick your document from 156 passport and visa specs across 150+ countries — US, UK, India, Japan, China, the Schengen countries and many more. Take a selfie with the front camera or upload an existing photo. The tool sets the crop to your document’s pixel size with a head-height guide. Where the document’s rules allow edits, it removes the background automatically; where they forbid edited photos (the US, UK, Canada, India and a few others), it crops and resizes only. Where the authority publishes file rules, the result checks the JPEG’s pixels, file size and format against them and links the official page. Download. Print the sheet at any drugstore that prints 4×6 photos, or upload the JPEG file directly to the digital application.
Zero install. Zero account. Your selfie never leaves the device — the background removal runs on your phone, not on someone else’s server. Same URL works on iPhone Safari, Android Chrome, an iPad, or a desktop browser.
A passport photo isn’t one photo. It’s many subtly-different photos depending on which government’s database it goes into.
A printed US passport photo is a 2 × 2 inch square with the head 25-35 mm tall; the digital photo for an online renewal must be a JPG, JPEG, PNG, HEIC or HEIF file between 54 KB and 10 MB. A US visa photo is a different file: a square JPEG of 600 to 1200 pixels, 240 kB or less. A UK digital passport photo must be at least 600 × 750 pixels and between 50 KB and 10 MB; the printed one is 35 × 45 mm with a cream or light grey background and the head 29-34 mm tall (GOV.UK). An Indian passport photo is 630 × 810 pixels with a white background, “unaltered by computer software”, and prints at 35 × 45 mm (passportindia.gov.in). An online form can reject a file that misses any of these.
A studio photographer knows these by memory for the country they handle most often. A generic camera app does not. A “free passport photo” tool that uploads to a server may know them or may not; you find out when the application is rejected, which costs you the application fee and the trip date.
What the ID Photo Maker tool does is bake the spec into the crop. Pick “US Passport” and the crop overlay locks to 600 × 600 with the head-height window shaded in the safe zone. Pick “UK Passport” and the crop becomes 630 × 810, above GOV.UK’s digital minimum, with a different head guide. Where the authority publishes a file-size range, the JPEG quality is tuned to land inside it, and the result lists each rule it checked with a link to the official page. The user doesn’t have to memorise the spec — the preset encodes it.
The hard part of an ID photo from a selfie is the background. Studios use a physical white sheet or a controlled lighting rig. At home you’re using whatever wall you happen to be standing next to, which is rarely the right colour and often has shadows.
The fix is portrait matting — a small machine-learning model that identifies the silhouette of a person against any background and produces an alpha mask so the background can be replaced. The model that powers Loft’s tool is U-2-Netp, a small publicly-released segmentation network of about 5 MB that runs in your browser through ONNX Runtime Web, in a background worker.
What that means in practice: the first time you open the tool, your browser downloads the 5 MB model. On a 4G connection that takes one or two seconds. The model is then cached in your browser, so every subsequent use is offline-capable. The inference itself runs locally, on WebAssembly. A typical phone processes a selfie in two or three seconds.
For documents whose rules say the photo must be unaltered — the US, UK, Canada, India, South Korea, New Zealand and Australia — the tool skips this step by default and only crops and resizes your photo, so shoot against a plain light wall.
The relevant point for ID photo specifically: your selfie does not leave your device. The model file does — once, at first run, from our R2 CDN — but no part of your photo is uploaded anywhere to be processed. The pillar at /docs/how-it-works/ covers the verification path; the short version is that you can open the browser’s Network tab and watch the tool run without seeing your photo in the request list.
The tool’s wizard walks through four steps. You don’t need to memorise them; the UI guides each one. But here’s what’s happening underneath.
Step 1 — pick the document. A scrolling list of 156 passport and visa specs across 150+ countries, searchable by country. Each preset sets the output pixels, the print size, a head-height guide and, where the authority publishes them, the file-size and format rules. The crop overlay updates as soon as you pick.
Step 2 — take a selfie or upload. On a phone, the front camera opens directly. There’s a safe-zone overlay showing where your head should sit. If you’ve already taken a photo (e.g. one on iCloud Photos or in your gallery), you can upload that instead. The tool accepts JPEG, PNG, WebP, and — importantly for iPhone — HEIC and HEIF input, so iPhone photos work without conversion.
Step 3 — crop and background. This is where the matting model runs, for documents that allow it. Your photo appears with the background already removed, sitting on top of a default solid colour; for a crop-only document it appears as you shot it. Pinch-zoom to position the head inside the safe-zone overlay. Pick a different background colour if the document requires one — white (most passports), off-white, light grey (UK, several EU), or light blue (some employee badges).
Step 4 — result. Download. Where your document has a published file-size range, the JPEG quality is tuned to land inside it. The result lists each file rule it checked against the official page, and what it does not check: head size, expression, lighting and background evenness. The print sheet places each copy at the document’s size in millimetres. The filename includes the preset name so you can find it later.
The whole thing is quick — the first run includes a model download; after that, subsequent uses are near-instant.
In January 2026 the US State Department updated its photo guidance to reject photos with AI-edited facial features. This followed a year of reports of submissions where the applicant’s face had been silently retouched by mobile apps.
For a US passport or visa the ID Photo Maker does not touch the photo by default: it crops and resizes only. Where it does replace a background, it only segments it — it does not retouch the face, the skin, the eyes, or anything else about the person. The U-2-Netp model is a segmentation network, not a generative or beautification network. It outputs an alpha mask for the silhouette; it doesn’t redraw any part of the subject.
This is a deliberate design choice. The US State Department photo requirements page is the authoritative source if you need to verify anything before submitting; this post and the tool are best-effort against published specs, not legal advice.
The tool works well for:
It does not work well for:
Three things make this work without anything to install.
The first is WebAssembly, which lets the browser run the matting model at speed comparable to a native binary. The model is the same .onnx file native iOS and Android apps would use; the browser just runs it in a different runtime.
The second is the Service Worker, which caches the tool and the model after first use so the second visit is offline-capable. You can pin the tool to your home screen via “Add to Home Screen” on iOS or “Install App” on Android, and after that it behaves like a native app — icon, standalone window, offline launch — without any actual native install.
The third is that browser hardware access — camera, file picker, gallery — has caught up with what native apps had access to. iOS Safari can use the front camera, Android Chrome can pick from Google Photos, both can read HEIC. The hardware no longer favours the native path the way it used to a few years ago.
The combination means a tool that genuinely had to be a native app five years ago can be a webpage today, and the user gets the same outcome with substantially less friction.
There are real cases where paying the $15 at the pharmacy or visiting the embassy-approved photographer is the right call:
For the routine adult passport or visa case covered by the preset catalog, the browser-based approach is the right shape. The studio is the right call for the edge cases.
Three quick checks before you commit a passport photo to the mail:
The dimensions match the consulate’s published spec. The preset should handle this, but eyeball it against the official guide linked from your application — a 600 × 600 file should look square, a 35 × 45 mm file should look slightly taller than wide.
The head height is inside the spec window. The safe-zone overlay during cropping is a guide: the tool does not measure your head, and it lets you download any crop. So check the downloaded JPEG against the head-size rule on the official page.
The file size is inside the published range. Where the authority publishes one, the tool tunes JPEG quality to land inside it and the result shows the check, but if you’ve re-saved the file through something else after download, the size may have shifted. Check the file size in your file manager before uploading.
That’s it. Selfie to printable JPEG, start to finish.
The ID Photo Maker is at
/tools/image-tools/id-photo-maker/. The privacy mechanism
that lets it run without uploading is covered in the
pillar at /docs/how-it-works/, section 4
specifically.
