Resize, crop, convert, watermark, and apply AI effects to any image — just by changing the URL. Powered by libvips for sub-50ms processing, cached at the edge for instant delivery.
Every transformation starts with an operation in the URL path that determines how the image is sized.
| Operation | URL Pattern | Description |
|---|---|---|
cover | /cover/600x400/ | Resize & crop to exact dimensions |
width | /width/800/ | Resize to width, height scales proportionally |
height | /height/600/ | Resize to height, width scales proportionally |
cdn | /cdn/ | Passthrough — auto-rotate, compress, serve |
Control exactly how the crop region is selected. Append a mode after cover:
Content-aware crop targeting the most visually interesting region — highest luminance and saturation.
Shannon entropy-based crop — focuses on the region with the most visual complexity.
ML-powered face detection — keeps detected faces in frame with configurable padding.
Fits image inside dimensions with padding. Combine with solid colours, gradients, or grid patterns.
Precise control with fp-x, fp-y, fp-z — position the crop center anywhere from 0.0 to 1.0.
10 directional modes: north, south, east, west, northeast, northwest, southeast, southwest, center, inside.
Serve the optimal format for every browser automatically. Image Foundry reads the Accept header and negotiates: AVIF → WebP → Original. The Vary: Accept header ensures CDN caches separate copies per format.
# Automatic — best format per browser
/my-images/width/800/format:auto/photo.jpg
# Explicit format
/my-images/width/800/format:avif,quality:60/photo.jpg
/my-images/width/800/format:webp,lossless:true/photo.jpg| Option | Values | Description |
|---|---|---|
format | auto, jpg, png, webp, avif, tiff, gif, svg | Output format |
quality | 1–100 | Compression quality (default 80) |
lossless | true | Lossless WebP/AVIF encoding |
progressive | true/false | Progressive JPEG/PNG |
dpr | 1–3 | Retina — multiplies dimensions (dpr:2 = 2× output) |
Composite logos, watermarks, and rendered text directly onto images — no pre-processing required.
# Watermark in the bottom-right corner
/my-images/width/800/wmk-path:%2Flogo.png,wmk-position:southeast,wmk-pad:15,wmk-opacity:80/photo.jpg
# Text overlay with drop shadow
/my-images/width/800/txt:Free%20Shipping,txt-size:48,txt-color:ffffff,txt-position:south,txt-shad:2/photo.jpg20 pre-installed Google Fonts (Roboto, Inter, Poppins, etc.) plus custom .ttf/.otf font support.
TensorFlow.js-powered automatic background removal. Replace with solid colour, add drop shadow, control feathering.
Detect faces and return bounding boxes as JSON. Apply privacy blur or pixelation automatically.
One parameter to blur or pixelate all detected faces — faceblur:10 or facepixel:15.
| Feature | Example |
|---|---|
| Region extraction | extract-top:100,extract-left:50,extract-width:400,extract-height:300 |
| Padding & borders | pad:20,fill-color:ffffff · border:5,ff0000 |
| Rounded corners & masks | corner-radius:20 · mask:ellipse |
| Image blending | blend-img:%2Ftexture.png,blend-mode:overlay,blend-alpha:50 |
| PDF page rendering | page:3 — renders specific page from PDF/multi-page TIFF |
| Colour palette extraction | palette:json,colors:8 or palette:css |
| SVG sanitization | Automatic script/event handler removal for uploaded SVGs |
| Animated GIF/WebP | animation:true (default) or animation:false to flatten |
No SDK required — just change the URL.