Quality
Controls the visual quality of the output JPEG. Higher values result in better quality and larger file sizes. Jpegli's perceptual model means that a quality of 90 is often visually identical to the original, but much smaller.
Chroma Subsampling
Reduce color detail (while keeping brightness sharp) to shrink files. This works because human eyes are less sensitive to color shifts than to brightness changes. Learn more ↗
Reduces color resolution by 75% (best general savings for photos).
Preserves full color resolution (prevents color bleed on sharp edges).
Reduces horizontal color resolution only.
Color Encoding
Choose the color model used during compression. The output is always a standard, compatible JPEG. XYB (perceptual) ↗, from the JPEG XL project, often gives smaller files for the same visual quality. YCbCr (legacy) ↗ matches classic JPEG math for maximum compatibility in strict, older workflows.
Progressive Mode
Progressive JPEG shows a quick full-image preview that sharpens as it downloads, which feels faster on slow networks.
Sequential loads from top to bottom and can decode slightly faster on older devices.
Learn more ↗
Default. More scans can improve perceived loading.
Fewer progressive stages.
Standard top-to-bottom decoding.
Adaptive Quantization
Smarter compression that spends more data (bits) on detailed, complex areas of the image and less on smooth, simple areas. This generally improves visual quality at a given file size. Highly recommended.
Transparency & Background
JPEG files cannot be transparent. When converting an image with transparent areas (like a PNG or SVG), this color will be used for the background. This is also known as the "matte" color. Learn more ↗
Grayscale
Remove color entirely to greatly reduce size (great for scans, B&W photos, line art).
Metadata
Choose which metadata to keep from the original file. Turning these off saves space and improves privacy.
Note: EXIF orientation is always applied to the image pixels to prevent rotation issues, then stripped. Not all source formats (like PDF or SVG) contain all types of metadata.
Download
When finished, automatically start a ZIP download.
Rasterization Defaults
Default settings for converting vector or document formats into pixels (a process called rasterization).
PDF & TIFF (multi-page)
Configure how documents with multiple pages, like PDFs and some TIFFs, are converted.
How to export multiple pages:
SVG rasterization
Control the output pixel dimensions when converting SVG images.
Used only when Width and Height are both 0. We scale so the longer side matches this value.
Non-zero Width/Height override “Longest side”. If one is 0, we compute it using the SVG’s aspect ratio.
Privacy
Contribute anonymous usage statistics to help improve the service. No images or personal data are ever sent.
Reset
If you get lost, reset back to safe defaults.