How it works
Choose a photo, logo or screenshot and its main colors are found with the median cut algorithm, which repeatedly splits the image's colors into groups of equal size. Each color shows its HEX code and how much of the image it covers, and the whole palette can be copied as CSS variables.
Image → colors grouped by similarity → average color of each group
Good uses
- Build a brand or website palette from a product photo.
- Match slide, poster or thumbnail colors to an image.
- Find the exact colors of a logo or illustration.
Tips
- Fewer colors give the dominant tones; more colors reveal accents.
- Transparent areas of PNG images are ignored.
- The image is analyzed on your device and never uploaded.
Standards and sources
- Color quantization uses the median cut algorithm (Heckbert, 1982)
- Colors are reported in sRGB as defined in CSS Color Module Level 4
Last reviewed October 7, 2026
Frequently asked questions
How are the main colors chosen?
The median cut algorithm repeatedly splits the image colors into groups and takes the average of each group, so frequent colors stand out.
How many colors should I extract?
Four to six give the overall mood; eight to twelve also catch smaller accents.
Does it work with logos and screenshots?
Yes. Flat images give very exact colors, and transparent PNG areas are ignored.
Is my image uploaded?
No. The image is resized and analyzed inside your browser and never leaves your device.
Can I use the colors in CSS?
Yes. Copy the CSS variables block, or tap a color to copy its HEX code.