Image Color Palette Extractor: Frequently Asked Questions

PNG, JPG/JPEG, WEBP, and GIF (first frame only) are supported. All processing happens in your browser using the Canvas API — your image is never uploaded to any server.

The algorithm extracts dominant colours, not every colour in the image. Very dark and very light pixels are excluded to prevent the palette from being dominated by shadows and highlights. The clustering averages nearby colours, so the palette colours may be slightly different from any individual pixel in the image.

The colours themselves (hex codes) are not copyrightable. You are free to use extracted colour palettes in any commercial project. The image you upload may be subject to copyright if it is not yours — but the extracted colours are just numbers.

The k-means algorithm uses random initial cluster positions, which can lead to slightly different final palettes across runs. The most dominant colours will consistently appear, but the order and exact values of secondary colours may vary. Rerunning the extraction is useful if the first result does not capture the colour you were looking for.

The tool samples the image at 100×100 pixels for processing speed regardless of the original file size, so there is no practical limit on image dimensions. Very large file sizes may take a moment to read into the browser, but processing itself is fast.

Yes — take a screenshot of any web page or UI (Command+Shift+4 on Mac, Snipping Tool on Windows), save as PNG, and upload. This is useful for analysing competitor colour palettes or auditing your own site's colour usage.