Squoosh
Drops an image into the browser, shows the compressed version beside the original, and never uploads the file — free and open source.
Quick facts
| Pricing | Free |
|---|---|
| Free tier | Yes — the whole tool is free, with no account and no paid tier |
| Platforms | Web |
| Developer | Google Chrome Labs |
| Website | https://squoosh.app/ |
What Squoosh is
Squoosh is a single-purpose image tool that runs entirely inside the browser tab. You drop, paste or pick an image, choose an output codec — MozJPEG, WebP, AVIF, OxiPNG, JPEG XL or QOI — and drag a quality slider while the screen shows the original and the result side by side, split by a movable line. The file size updates as you change settings, so the trade between weight and visible damage is decided by looking rather than by guessing. Resize and palette reduction sit in the same panel, which covers most of what a page image needs before upload. Because the work happens locally in WebAssembly, the file never leaves the device, and the app keeps working once loaded, even offline. It is open source, published by Google Chrome Labs, and free. The one real limit is throughput: it handles one image at a time, so a folder of two hundred files is a job for the command line version, not this page.
Best for
- Cutting a hero image down to size before it goes on a page
- Comparing WebP and AVIF output at the same visual quality
- Compressing a photo you would rather not hand to a server
Not ideal for
- Batch jobs across a folder of images — the web page takes one at a time
- Pipeline work that has to run unattended on every build
Pros
- Images stay on the device; nothing is uploaded
- Side-by-side comparison with live file size while you tune quality
- Free, open source, no account and no watermark
Cons
- One image per session, so bulk work means switching to the CLI
- Every codec exposes its own parameters, which rewards knowing what they do
- No history or library: save the file or lose the settings
Pricing in detail
Free, with nothing to buy. There is no account, no subscription, no paid tier and no file-count limit — the page never asks for a card. It is an open-source project published by Google Chrome Labs, so the source can be read, self-hosted or modified under its licence. What it costs you is bandwidth to load it the first time and processor time to encode, which scales with image size and codec; AVIF and JPEG XL are noticeably slower than MozJPEG on a large photo. A paid folder-based compressor only makes sense if you routinely handle batches, since this app works one image at a time.