Differences

Comparison with Other Libraries

Modern Colorthief differs from similar libraries in several key ways.

vs fast-colorthief

Feature	Detail
Architecture Support	Broader support via PyO3 (vs pybind11)
NumPy Dependency	No hard dependency on NumPy
Codebase	Simpler, more maintainable Rust code vs C++ codebase
Build Tooling	Automated via Maturin and GitHub Actions
Package Size	500-700 KB vs 52-60 KB for fast-colorthief

vs color-thief-py

Feature	Detail
Execution Speed	Nearly 100x faster
Pillow Dependency	No hard dependency on Pillow
Python Compatibility	Fully compatible with modern Python versions

Achieving Parity with colorthief (Python)

If you need output that matches the original colorthief library:

1. Load the image with Pillow.

2. Save the loaded image to a BytesIO object.

3. Pass the BytesIO object to modern_colorthief.

import io
from PIL import Image
import modern_colorthief

path = "photo.jpg"
img = Image.open(path, mode="r")

image_bytes = io.BytesIO()
img.save(image_bytes, format="PNG")
image_bytes.seek(0)

dominant_color = modern_colorthief.get_color(image_bytes)
dominant_palette = modern_colorthief.get_palette(image_bytes)

Achieving Parity with fast-colorthief (C++)

The same approach works for matching fast-colorthief output:

import io
from PIL import Image
import modern_colorthief

path = "photo.jpg"
img = Image.open(path, mode="r")

image_bytes = io.BytesIO()
img.save(image_bytes, format="PNG")
image_bytes.seek(0)

dominant_color = modern_colorthief.get_color(image_bytes)
dominant_palette = modern_colorthief.get_palette(image_bytes)

Note

The parity technique works because Pillow normalizes pixel data across image formats, ensuring consistent input to the quantization algorithm.

See also

API Differences – Detailed output comparison with color-thief.

Benchmarks – Speed comparison across implementations.