Shrink your files, keep the magic. Premium file compression for creators who care about quality.
Compressly is a privacy-first file compression tool that lets you reduce file sizes exactly the way you want — either by choosing a compression level or by targeting a specific output file size. All compression happens locally in your browser or on the edge. Your files are never uploaded to or stored on any cloud server, so your data stays yours.
- 🎚️ Compression levels and exact targets — Pick a level, or set a target size per file. One-click presets for WhatsApp (16 MB), Discord (10 MB), email (25 MB) and form uploads (200 KB / 100 KB).
- 🖼️ Images — JPG, PNG, WebP, BMP and iPhone HEIC in; same format, JPG, WebP or AVIF out. Optional max dimensions. Before/after comparison slider.
- 📄 PDFs — "Keep text selectable" re-compresses only the images inside, or "Smallest size" turns pages into images.
- 🎬 Video and audio — MP4, WebM or GIF output, resolution, remove audio, trim. Audio becomes MP3.
- 🎞️ GIF and SVG — Animated GIFs keep every frame (gifsicle); SVGs are optimized with SVGO.
- 🔒 Private by design — Everything runs in your browser. Photo location and camera data is removed unless you choose to keep it.
- 📱 Installable app — Works offline once loaded; on Android, share files to Compressly from any app.
- ✨ Comfortable to use — Paste or drop whole folders, cancel any file, per-file options, remembered settings, dark mode.
| Layer | Technology |
|---|---|
| Frontend | TypeScript, Vite, HTML/CSS |
| API / Backend | Cloudflare Workers (via worker/) |
| Deployment | Cloudflare Workers (static assets) |
| Config | wrangler.toml for Cloudflare settings |
- Node.js (v18 or later recommended)
- npm
- A Cloudflare account (for deployment)
# 1. Clone the repository
git clone https://github.com/HapticHash/compressly.git
cd compressly
# 2. Install dependencies
npm install
# 3. Start the development server
npm run devThe app will be available at http://localhost:3000 by default.
npm run lint # type-check
npm test # unit tests
npm run test:e2e # browser tests against the production build (Playwright)Run npx playwright install chromium once before the browser tests.
npm run buildThe output will be in the dist/ folder.
Compressly deploys as a Cloudflare Worker with static assets: the Vite build in dist/ is served as static files, and /api/* requests are handled by the Worker in worker/. Configuration lives in wrangler.toml.
npm run build
npx wrangler deployWith Workers Builds (Git integration), use npm run build as the build command and npx wrangler deploy (or npx wrangler versions upload) as the deploy command.
Use npx wrangler deploy rather than npx wrangler versions upload: the stats counter is a Durable Object, and the first deploy has to apply its migration, which version uploads can't do.
Optional build-time environment variables:
| Variable | Purpose |
|---|---|
VITE_SITE_URL |
Public URL used in canonical links, og:image and sitemap.xml (default: https://compressly.harshitks2203.workers.dev) |
CF_ANALYTICS_TOKEN |
Enables Cloudflare Web Analytics |
To run the production setup locally, including the API and stats counter:
npm run build
npx wrangler devcompressly/
├── docs/
│ └── screenshots/ # README images (not deployed)
├── worker/ # Cloudflare Worker: /api/* routes, stats Durable Object
├── tests/e2e/ # Playwright browser tests
├── scripts/ # Generates the PWA icons and og-image.png
├── public/ # Static assets
├── src/
│ ├── components/ # UI components
│ ├── hooks/ # Settings and theme state
│ ├── lib/ # Compression logic, loaded on demand per file type
│ └── sw.ts # Service worker (offline, caching, share target)
├── index.html # App entry point
├── vite.config.ts # Vite configuration
├── wrangler.toml # Cloudflare Worker config
└── package.json
Compressly is built with privacy as a core principle:
- Files never leave your device — All compression runs in the browser (WebAssembly and web workers). No file is uploaded, stored or logged.
- Metadata removed by default — Re-encoded photos drop EXIF data such as GPS location, unless "Keep photo metadata" is on.
- Minimal network use — The only requests with user activity are an anonymous count of files and bytes saved for the public stats. Fonts and libraries are self-hosted; the FFmpeg core (~31 MB, too large for Cloudflare's 25 MiB file limit) is fetched from unpkg only when you add a video or audio file, then cached.
- Optional analytics — If
CF_ANALYTICS_TOKENis set at build time, Cloudflare Web Analytics counts page views. It uses no cookies and never sees your files.
You can verify this by inspecting the source code in worker/ and src/.
Contributions are welcome! Here's how to get started:
- Fork the repository
- Create a new branch:
git checkout -b feature/your-feature-name - Make your changes and commit:
git commit -m "feat: add your feature" - Push to your fork:
git push origin feature/your-feature-name - Open a Pull Request
Please make sure your code follows the existing TypeScript conventions and the project builds without errors before submitting.
This project is licensed under the MIT License.
Built with ❤️ by HapticHash. If you find this useful, consider leaving a ⭐ on the repo!