Color contrast that works for everyone

CM-Colors automatically fixes your colors so they are easy to read. It finds similar colors that work for everyone, so you don't have to guess.

"Fortunately, there's a tool that's just come out that's going to change your life: CM-Colors. You give it your colors, and it automatically adjusts them so they're accessible, all while changing the shades as little as possible to keep your design intact."

— Korben.info

The percentage shows how much easier the text is to read:

Spending too much time adjusting colors? CM-Colors handles it for you. It automatically finds colors that look like your brand but are readable for all your visitors.

What it does:

• Fixes colors automatically: No more manual tweaking.
• Works everywhere: Fix Python code or CSS files.
• Keeps your style: Changes colors as little as possible.
• Saves you time: Fix your whole project in seconds.

pip install cm-colors

from cm_colors import ColorPair

# Your colors
pair = ColorPair("#999999", "#ffffff")

# Fix them and preview in the terminal
fixed_color, success = pair.make_readable(show=True)

print(f"Use {fixed_color} instead of #999999")
# Output: Use #8e8e8e instead of #999999

from cm_colors import make_readable_bulk

my_colors = [
    ("#777", "#fff"),
    ("#888", "#000"),
]

results = make_readable_bulk(my_colors)

for color, status in results:
    print(f"{color} is {status}")

# check a pair
cm-colors contrast '#777777' '#ffffff' --json

# fix a pair
cm-colors fix '#777777' '#ffffff' --json

# fix a CSS file
cm-colors fix styles.css

cm-colors fix styles.css creates styles_cm.css with readable colors so you can preview before touching your original.

For more details, see the full documentation.

The optimization algorithm is described in two peer-reviewed papers:

• Context-Adaptive Color Optimization for Web Accessibility (arXiv:2512.07623) — context-adaptive constraint strategies achieving 93.68% success across random color pairs and 100% on pairs with initial contrast ratio > 2.0, tested on 10,000 programmatically generated branding-representative pairs.

• Perceptually-Minimal Color Optimization for Web Accessibility (arXiv:2512.05067) — the multi-phase OKLCH optimization approach that minimizes perceptual change (CIEDE2000) while meeting WCAG contrast requirements.

GNU General Public License v3.0

Found a problem? Let us know.

Stop guessing, start building.