---
id: rawpy
version: "0.27.0"
license: MIT
license_treatment: permissive
maintenance: active
---

# rawpy — RAW image processing for Python, a wrapper for libraw

License: permissive · Maintenance: active · Downloads: 532.0K/mo

## What it is and what it does

rawpy is a Python wrapper around the LibRaw C++ library that handles the low-level decoding of RAW image files from digital cameras. It lets you load RAW files (NEF, CR2, ARW, DNG, and many others), extract metadata and embedded thumbnails, and apply postprocessing such as demosaicing, white balance, gamma correction, and output bit depth control. The package also includes utilities in the `rawpy.enhance` module for detecting and repairing hot and dead pixels across multiple images.

The library is built as precompiled wheels for Python 3.9 through 3.14 on Linux, macOS, and Windows, with optional features like LCMS color management and OpenMP acceleration available on most platforms. It depends only on numpy and is designed for straightforward use in image processing pipelines, from simple one-line postprocessing to advanced workflows involving pixel repair and batch processing.

Use it for:

- Load and postprocess RAW files from digital cameras with custom white balance, gamma, and bit depth settings.
- Extract and save embedded thumbnail images from RAW files without full decoding.
- Identify and repair hot/dead pixels across a series of RAW images using statistical analysis.
- Build batch RAW processing pipelines that apply consistent postprocessing parameters to multiple images.
- Access raw sensor data and metadata for scientific or forensic image analysis.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

rawpy wraps LibRaw to load, decode, and postprocess RAW image files from digital cameras, with built-in functions for finding and repairing hot and dead pixels.

Yes. rawpy is actively maintained, has no known vulnerabilities, and provides the most straightforward Python interface to LibRaw for RAW image processing. Install it if you need to work with RAW files from digital cameras. Be aware of the medium install friction (compiled wheels) and the Linux multiprocessing fork() caveat documented in the FAQ.

## Install

pip install rawpy
uv add rawpy
poetry add rawpy

## Installing rawpy

Before you install: Medium install friction due to compiled wheels for multiple Python versions (3.9–3.14) and platforms. Package is actively maintained with recent commits and no known vulnerabilities. Depends only on numpy, keeping the dependency tree lean.

License in practice: MIT license is permissive and imposes no restrictions on use, modification, or distribution. Note that GPL demosaic packs are intentionally excluded from wheels due to license incompatibility with MIT.

Quickstart:

pip install rawpy

import rawpy
import numpy

with rawpy.imread('image.nef') as raw:
    rgb = raw.postprocess()

LibRaw library must be available on the system; on Linux, manual installation from source may be required if the system package is outdated or missing.

Verify before relying:

- Whether all supported camera models and RAW formats are documented or queryable at runtime.
- Performance characteristics when processing large batches of RAW files or very high-resolution images.
- Stability and thread-safety guarantees beyond the documented multiprocessing fork() caveat.

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 532.0K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

raw image processing python, libraw wrapper, camera raw file reader, raw postprocessing, hot pixel repair, raw image decoder, digital camera raw support, image-processing, raw-photography, libraw-binding

[View on SkillFed](https://skillfed.io/packages/rawpy) · [View on PyPI](https://pypi.org/project/rawpy/)