Benchmarks¶

Overview

This page compares the execution time of modern_colorthief (Rust-based) against the original colorthief (pure Python) and fast_colorthief (C++ based).

Benchmark Script

You can reproduce the benchmarks using the following script:

benchmark_colorthief.py¶

import os
import timeit
from pathlib import Path

from colorthief import ColorThief
from fast_colorthief import get_dominant_color, get_palette as f_get_palette
from modern_colorthief import get_color, get_palette

BASE_DIR = Path(__file__).resolve().parent
path = os.path.join(BASE_DIR, "test.jpg")

# --- Extracting Color ---
start_time = timeit.default_timer()
ColorThief(path).get_color()
py_color_time = timeit.default_timer() - start_time

start_time = timeit.default_timer()
get_dominant_color(path, 10)
cpp_color_time = timeit.default_timer() - start_time

start_time = timeit.default_timer()
get_color(path)
rust_color_time = timeit.default_timer() - start_time

# --- Extracting Palette ---
start_time = timeit.default_timer()
ColorThief(path).get_palette()
py_palette_time = timeit.default_timer() - start_time

start_time = timeit.default_timer()
f_get_palette(path)
cpp_palette_time = timeit.default_timer() - start_time

start_time = timeit.default_timer()
get_palette(path)
rust_palette_time = timeit.default_timer() - start_time

print(f"Python color:  {py_color_time:.6f}s")
print(f"C++ color:     {cpp_color_time:.6f}s")
print(f"Rust color:    {rust_color_time:.6f}s")
print(f"Python palette: {py_palette_time:.6f}s")
print(f"C++ palette:    {cpp_palette_time:.6f}s")
print(f"Rust palette:   {rust_palette_time:.6f}s")

Results

On a sample image, the execution times are approximately as follows:

Performance Comparison¶
Task	Python (colorthief)	C++ (fast_colorthief)	Rust (modern_colorthief)
Extracting Color	0.219895 s	0.021180 s	0.019645 s
Extracting Palette	0.202956 s	0.023626 s	0.018661 s

Warning

Benchmark results may vary depending on the image size, color complexity, and hardware. Use the script above to test on your own workloads.