Mochify Frontend
Privacy-First • Hardware-Hardened • Green C++ Engine
Mochify is a high-performance media processing utility for images, PDFs and video. Unlike traditional cloud converters that buffer uploads to disk, Mochify uses a stateless C++ engine to ensure your pixels never touch permanent storage.
Launch App →
What it does
Images
Convert and compress between JPG, PNG, WebP, AVIF and JPEG XL, with HEIC/HEIF/HIF and SVG accepted as input. Resize, rotate, crop, strip EXIF, and remove backgrounds. Saliency-guided smart compression softens the background while leaving the subject untouched, and smart crop centres on the detected subject instead of the geometric middle.
Describe what you want in plain English ("convert to WebP and AVIF at 1200px, strip the metadata") and Magic Flow maps it to parameters, or use the classic form if you would rather set them yourself.
PDFs
A full PDF toolkit at POST /v1/pdf:
| Operation | What it does |
|---|
optimize | Recompress the images inside a PDF, in place. Text, fonts, vector art, links and layout are untouched, so the document stays searchable. |
extract | Pull the embedded images out at the resolution they were stored at. Not a render of each page: the actual pictures somebody placed in the document. |
rasterize | Render every page to an image. |
split | Explode a PDF into one single-page PDF per page. |
create | Build a PDF from images, one page per image. |
optimize targets resolution against how large each image is actually drawn on the page, read from the content stream, rather than how many pixels it happens to store. A 4000px photo dropped into a two-inch box is 2000 DPI of waste; the same photo full-bleed on A4 is not, and only the drawn size tells them apart. Images it cannot confidently handle are passed through byte-for-byte rather than dropped, and it never returns a file larger than the one you gave it.
Video and audio
MP4, WebM, MKV and MOV conversion plus audio extraction, running entirely in your browser via MediaBunny. This is the one part of Mochify that is genuinely client-side: the file never leaves your machine, and the server never sees it.
Save to your own storage
Results can be written straight into your own S3 / R2 / S3-compatible bucket or Google Drive instead of downloading. The bytes go from Mochify to your storage directly rather than back through the browser.
Privacy Model
This frontend is open-source so you can verify exactly how your data is handled before it leaves your browser.
- Public UI (SvelteKit + Cloudflare): Auditable code. No hidden trackers. No third-party ad networks.
- Private Vault (C++): A proprietary, hardware-locked engine running on native Linux kernel primitives to guarantee data volatility.
Security Hardening
- Volatile RAM only: Temporary file creation is disabled and container swap is off — images are buffered in volatile RAM and wiped on request completion.
- Zero-buffer streaming: Data streams directly from the TLS connection into the C++ process; no intermediate disk writes.
- Strict CSP: A rigid
connect-src policy prevents the browser from sending data to any domain outside our verified API.
Analytics
We use self-hosted Umami — cookie-free, GDPR-compliant, and anonymized. No data is shared with Google, Facebook, or any ad networks.
AI & MCP Integration
Mochify exposes a hosted Model Context Protocol server at mcp.mochify.app, letting AI agents compress and convert images directly without any manual tooling.
Add to Claude.ai or any MCP-compatible client:
https://mcp.mochify.app
Or install via Smithery:
Auth is handled via OAuth — no API key configuration required. The squish tool accepts a public image URL plus optional format, dimensions, and quality parameters.
For CLI and local MCP use, see mochify-cli.
REST API: full parameter reference, response headers and cURL / JavaScript / Python examples at mochify.app/docs. A machine-readable summary for agents lives at mochify.app/llms.txt.
Performance
Built with native C++ and libvips, ditching heavy interpreted runtimes for real gains:
- Energy efficient: Native code uses a fraction of the electricity per megapixel compared to Python/Node.js-based APIs.
- Low latency: Average processing time ~822ms, with disk I/O eliminated as a bottleneck.
Developing
Stack: Svelte 5, TailwindCSS v4, Cloudflare Pages adapter.
Prerequisites: Node.js 20+
git clone https://github.com/tliesnham/mochify-frontend.git
cd mochify-frontend
npm install
Copy .env.example to .env:
PUBLIC_API_URL=http://localhost
Note: The production API (api.mochify.app) enforces strict CORS/Referrer checks and will reject requests from localhost. Point PUBLIC_API_URL at a local instance of mochify-core for development.
npm run dev # start dev server
npm run build # production build
npm run check # type-check
npm run lint # lint + format check
Self-Hosting
The core engine is available as a hardened Docker image with the same RAM-only, zero-persistence config used in production. Multi-arch: amd64 and arm64 (Apple Silicon / AWS Graviton).
Docker Hub: mochify/mochify-lite
docker pull mochify/mochify-lite:latest
