Pulling a color palette out of any image
Upload an image and this tool samples its pixels, groups similar colors together, and surfaces the most common ones as a palette of hex swatches - click any swatch to copy its hex code. Useful for pulling brand colors from a logo, matching a design to a photo, or just exploring an image's color scheme. Everything runs locally in your browser; your image is never uploaded.
Why similar colors get grouped into buckets before counting
A real photo can contain thousands of subtly different shades of what looks like one color to the human eye - counting every exact pixel value separately would produce a palette cluttered with near-duplicate hex codes. Grouping nearby color values into buckets before counting collapses those near-duplicates into one representative swatch, which is what makes the result a usable palette rather than a long, redundant list.
Why transparent pixels are deliberately excluded from the count
A transparent or mostly-transparent pixel doesn't actually represent a visible color in the image - letting those pixels into the count would skew the palette toward whatever default color sits underneath transparent areas, which is why pixels below a transparency threshold are skipped entirely before colors are tallied.
Why analyzing a smaller version of the image barely affects the result
Dominant colors are, by definition, colors that occupy a large share of the image - a smaller sample still captures that same large-scale color distribution reliably, which is why reducing the image for faster processing has minimal effect on which colors come out as "dominant" for most photos and graphics.