Material Palette Generator

A Python port of the Official Material Palette Generator.
This project reverse-engineers and refactors Google’s original algorithm to generate palette (extracted by edelstone/material-palette-generator) into clean, reusable Python code.

Features

• Generate full Material Design palettes from a single base hex color
• Produces primary, complementary, analogous and triadic variations
• Shows palette preview in image (pillow library required)
• CLI support (mpg command)
• MIT Licensed
• No dependencies (pure Python)

Installation

Clone the repository:

git clone https://github.com/blacksteel3/material-palette-generator.git
cd material-palette-generator

Or install via pip:

pip install material-palettes-generator

Usage

import material_palette_generator as mpg

# generate palette from generic method
palettes = mpg.get_palettes("#E91E63")  # optional `types` parameter to specify color types

print(palettes['Primary'])
print(palettes['Complementary'])
print(palettes['Analogous-1'])
print(palettes['Analogous-2'])
print(palettes['Triadic-1'])
print(palettes['Triadic-2'])

# or use specific method for specific type
primary_palette = mpg.get_primary_palette("#E91E63")
print(primary_palette)

# preview any palette (`pillow` library must be installed)
mpg.preview_palettes(primary_palette)

Command Line Interface (CLI)

After installing, you can use the mpg command directly from your terminal:

# Generate the primary palette
mpg "#3f51b5"

# Output as JSON
mpg "#3f51b5" --json

# Output as CSS
mpg "#3f51b5" --css

# Generate specific types
mpg "#3f51b5" --primary
mpg "#3f51b5" --complementary
mpg "#3f51b5" --analogous
mpg "#3f51b5" --triadic

# Combine multiple flags
mpg "#3f51b5" -p -t --json

# Or get all palettes
mpg "#3f51b5" --all

# Preview palette
mpg "#3f51b5" --show

# See all commands
mpg --help

Credits

• Google — for creating the Material Design color algorithm
• Michael Edelstone — for extracting the obfuscated JavaScript runtime code
• Black Steel — for reverse-engineering, renaming, and porting the algorithm to Python