Skip to content
THE IMAGE LIBRARY FOR PYTHON

Make every pixel count.

Familiar Python image workflows, powered by Rust. Open, transform, and save images with a focused Pillow-shaped API.

Get started Explore the API

Performance compared with Pillow

Resize a 1920 × 1280 RGB image to 960 × 640. In this run, Blanket was 2.6–7.5× faster across all six resize filters.

Resize filter Blanket Pillow Speedup
Nearest 0.085 ms 0.259 ms 3.05×
Box 2.139 ms 5.537 ms 2.59×
Bilinear 2.228 ms 8.521 ms 3.83×
Hamming 2.293 ms 8.550 ms 3.73×
Bicubic 2.062 ms 15.370 ms 7.46×
Lanczos 2.849 ms 21.113 ms 7.41×

Median of 50 runs after 3 warmups, using identical synthetic in-memory inputs. Apple M1 Max, macOS 27.0 arm64, Python 3.14.7, Blanket 0.0.1 (release build), Pillow 12.3.0. Measured September 23, 2026. Speedup is Pillow time divided by Blanket time; higher is better. Results vary by hardware, image size, and operation.

Methodology, raw results, and reproduction commands.

BUILT FOR REAL WORK

Everything you need to move from source to output.

01 / START HERE

Work the way you know

Open images, apply transforms, and save the result with an API that feels familiar.

Read the usage guide → 02 / FORMATS & PRECISION

More room for your images

Choose a format and preserve higher bit depths where supported.

Explore formats → 03 / PIXEL TOOLS

Make the next edit

Resize, crop, colorize, enhance, and filter with native kernels behind a Python interface.

Browse operations →

A small example, start to finish

from blanket import Image, ImageOps

with Image.open("photo.jpg") as image:
    upright = ImageOps.exif_transpose(image)
    thumbnail = ImageOps.fit(upright, (256, 256), method=Image.Resampling.LANCZOS)
    thumbnail.save("thumbnail.webp", quality=85)

Blanket is alpha software. It supports L, RGB, and RGBA images, with limited indexed P support. See Pillow interoperability for integration with Pillow.

Find your next step