Automatic palette plus exact evidence
Dominant colors reveal the visual language of a photo, but they cannot replace exact sampling. This tool keeps both: a ranked palette for direction and a point picker for the precise pixel you need.
Extract the dominant palette, inspect an exact pixel and leave with production-ready color values — not just a row of pretty swatches.
Click or tap any point for its exact color.
Select a swatch to inspect it. Edit the selected color or add a sampled pixel.
The image is decoded and analyzed in browser memory. No request contains your file.
Dominant colors reveal the visual language of a photo, but they cannot replace exact sampling. This tool keeps both: a ranked palette for direction and a point picker for the precise pixel you need.
Every selected color includes HEX, RGB, HSL and its contrast against white and black. The AA result uses the WCAG 4.5:1 threshold for normal text, so the recommendation is measurable rather than decorative.
Copy a CSS custom-property block, a plain HEX list, or download JSON and a shareable PNG palette card. Everything is generated from the current edited palette.
Pulling colours out of an image is usually a step toward matching something: a site theme to a product photo, a slide deck to a logo, an illustration to a brand that was handed to you as a JPEG rather than as hex values. The extracted palette is a starting point for that work, and the gap between "these are the colours in the picture" and "these are the colours to use" is where the judgement lives.
A JPEG has been through lossy compression, which shifts colour slightly, especially at edges. A screenshot has been through the display pipeline. If a brand has published hex values, those beat anything sampled from an image of them.
Frequency-based extraction reports what covers the most area, which in a product shot is the seamless backdrop rather than the product. Read the palette as a list of candidates and pick by role — background, accent, text — rather than taking the top entry.
Two colours pulled from the same photograph are often close in lightness, which is exactly the condition that fails contrast requirements for text. Check any pairing you intend to use for body copy against a contrast ratio before committing to it.