The anatomy of a transform URL
Every mode explained
Resize ?r=...
scaleProportional scale to fit width/height. Preserves aspect ratio.
fitFit within the bounding box, letterboxed if needed.
fillFill the box, cropping to maintain aspect ratio.
cropHard crop to exact dimensions.
thumbSmart thumbnail optimised for face/object detection.
Gravity ?g=...
autoSmart gravity. Tuzzle picks the most visually interesting region.
faceCenter on a detected face. Falls back to auto if no face found.
facesFrame to include all detected faces.
eyesZoom to eye level for portraits.
centerCrop from the center. Classic fallback.
north/south/east/westCrop from a fixed compass direction.
Format ?f=...
webpLossy or lossless. Best for web. ~30% smaller than JPEG.
avifAV1 Image Format. Best compression. ~50% smaller than JPEG.
jpegUniversal compatibility. Best for photos.
pngLossless. Best for graphics with transparency.
gifAnimated GIFs preserved. Convert video frames to GIF.
tiffHigh-fidelity archival. For print and pro workflows.
Full parameter reference
| Param | Type | Description | Example |
|---|---|---|---|
w | number | Output width in pixels | w=800 |
h | number | Output height in pixels | h=600 |
f | enum | Output format: jpeg | png | webp | avif | gif | tiff | f=webp |
q | number | Quality 1–100 (lossy formats only) | q=85 |
r | enum | Resize mode: scale | fit | fill | crop | thumb | r=fill |
g | enum | Gravity: auto | face | faces | eyes | center | north | south | east | west | g=face |