AZ Tools

Color Name Finder

Color

Given any color (hex, RGB, or color picker), find the top 8 closest named colors from three databases: CSS 4 named colors (148 entries, what you can write directly in CSS like `cornflowerblue`), the default Tailwind v3 palette (110+ entries from `slate-50` to `rose-700`), and the XKCD color survey (the most common color names from a famous 222k-respondent study). Distance is measured in OKLab — a perceptually uniform color space — so the rankings actually match how a human eye sees similarity, unlike RGB Euclidean distance which over-weights luminance changes.

Closest matches
  • dodgerblue
    #1e90ff · ΔE 0.0363
  • royalblue
    #4169e1 · ΔE 0.0668
  • cornflowerblue
    #6495ed · ΔE 0.0697
  • mediumslateblue
    #7b68ee · ΔE 0.0871
  • steelblue
    #4682b4 · ΔE 0.1011
  • mediumpurple
    #9370db · ΔE 0.1131
  • slateblue
    #6a5acd · ΔE 0.1140
  • lightslategray
    #778899 · ΔE 0.1564

Distance is OKLab Euclidean — perceptually uniform. Smaller is closer.

How to use

  1. Enter a hex (`#3b82f6`, `3b82f6`, `#36e`), `rgb(59, 130, 246)`, or use the color picker.
  2. Switch between CSS, Tailwind, and XKCD databases.
  3. Copy a name to use directly in your CSS or Tailwind config.

Frequently asked questions

Why OKLab instead of RGB distance?
Pure RGB distance gives misleading rankings — a 10% brightness change in dark colors counts the same as in bright colors, but our eyes are much more sensitive to mid-tones. OKLab (released by Björn Ottosson in 2020) is calibrated for perceptual uniformity, so 'closest' really means 'looks most similar'.
What's the XKCD color survey?
In 2010, Randall Munroe ran a survey asking 222,500 people to name 5 random colors each. The most-agreed-upon names became a useful palette — sometimes used in data visualization libraries like matplotlib. Includes folksy names like 'puke green' and 'baby poop brown' (not in our top-60 subset).
Why only 8 results?
Beyond rank 8, distances grow large enough that 'closest' becomes meaningless. If you need a full sorted list, the algorithm is just OKLab Euclidean distance — easy to reproduce.

Related tools