How to Convert HEIC to JPG Without Uploading to Any Server

The problem with most converters

Search for “convert HEIC to JPG” and you get CloudConvert, iLoveIMG, Convertio, and a dozen similar sites. Every one of them uploads your file to a server, converts it there, and sends the result back. Your iPhone photos — trips, family events, medical documents, passport scans — travel to a company you have never heard of and sit there until their deletion timer runs out (usually 24–48 hours).

This was not a deliberate privacy choice by those companies. It was a technical constraint: HEIC decoding requires a specific library (libheif) that browsers do not include natively. Running the conversion on a server was the only practical option — until WebAssembly.

How WebAssembly changes this

WebAssembly (WASM) lets you compile C and C++ programs into a binary format that browsers can run at near-native speed. libheif — written in C++ — can be compiled to WASM and shipped as part of a web application. That is exactly what LocalJPG does, using the jsquash library to bridge C codecs to JavaScript.

The conversion pipeline runs entirely inside a Web Worker in your browser tab:

1. Your .heic file is read into memory in the Worker
2. libheif WASM decodes it to raw RGBA pixel data
3. MozJPEG WASM encodes the pixels to JPG at quality 85
4. EXIF metadata is injected back into the output file
5. The JPG is transferred back to the main thread as a Transferable (zero-copy)

Nothing in this pipeline requires a network request. The WASM binary is downloaded once on first visit and cached by the browser. After that, conversion works completely offline.

Verify it yourself

You do not have to take this on faith. Here is how to confirm that no image data leaves your device:

The NetworkCounter widget in the top right of LocalJPG shows this live: “image bytes sent to server: 0”. CloudConvert, iLoveIMG, and Convertio cannot honestly show that counter, because they do upload. LocalJPG can — and does.

What gets preserved

LocalJPG uses MozJPEG at quality 85 with 4:4:4 chroma subsampling. This preserves colour accuracy in photos with rich reds and skin tones — common with iPhone photos — at the cost of a slightly larger file than 4:2:0 subsampling would produce. EXIF metadata (date taken, GPS location, camera model, orientation) is injected into the output JPG via raw APP1 bytes.

Common questions