Color Palette Extractor
Upload an image and extract its most dominant colors as a palette of copyable hex and RGB swatches.
Input
JPG, PNG, GIF or WebP. Processed entirely in your browser — your image is never uploaded.
How many dominant colors to pull from the image. Range 3-10.
Palette
Upload an image and adjust the color count to extract its dominant palette. Everything runs in your browser — your image is never uploaded.
Guides
Upload any photo, illustration, or screenshot and this tool pulls out its most dominant colors, showing each one as a swatch with a ready-to-copy hex and RGB value. It's built for designers and developers who need a quick, accurate color palette from a reference image — a product photo, a brand logo, a piece of art, a screenshot of a UI you like — without opening an image editor.
How it works
- Upload an image (JPG, PNG, GIF or WebP).
- Drag the Number of colors slider (3-10) to choose how many swatches to extract.
- Each swatch shows its share of the image (the percentage of sampled pixels closest to that color), plus its hex and RGB codes — click either to copy it.
- Use Copy all hex codes to grab the whole palette as a newline-separated list, or Download palette PNG to save a single image with every swatch and its hex code — handy for dropping straight into a mood board or design file.
How the colors are chosen
The tool uses median-cut quantization, the same technique used by many professional color-extraction tools. Instead of just counting which exact pixel value shows up most often (which tends to produce a handful of near-identical, noisy shades on photos with gradients, shadows, or JPEG compression artifacts), it repeatedly splits the image's full cloud of colors along whichever channel — red, green, or blue — has the widest spread, then averages each resulting group into one representative color. The result is a small set of colors that actually represent distinct regions of the image, rather than whichever slightly-different shade happened to repeat the most.
Transparent pixels are ignored, and large images are downsampled before analysis so extraction stays fast without changing which colors dominate.
Is my image uploaded anywhere?
No. Decoding, sampling, and quantization all happen locally in your browser using the Canvas API — the image never leaves your device and nothing is sent to a server.
Why don't the swatches exactly match colors I picked out by eye?
Median-cut groups similar shades together and reports each group's average, so a swatch represents the center of a color region rather than any single pixel. For a photo with soft gradients or lots of texture, that average is usually a better summary of "the color used" than any one exact pixel value would be.
What's a good number of colors to extract?
3-5 works well for a simple logo or flat illustration with a few clear colors. 6-10 suits photos or textured images where you want to capture secondary and accent tones alongside the dominant background.