pylibraft-cu12
RAFT: Reusable Algorithms Functions and other Tools
Decision gist · record as of 2026-08-14
Yes, if you have a CUDA 12 GPU and need low-level accelerated primitives for algorithm development. The active maintenance, permissive Apache-2.0 license, and interoperability with numpy and other GPU libraries make it a solid foundation for GPU-accelerated workflows. Medium install friction is acceptable for the performance and code-reuse benefits. Not suitable if you lack GPU hardware or need high-level data science tools.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires NVIDIA GPU with CUDA 12 support and CUDA Toolkit installed; Python 3.11 or later.
- Medium install friction due to CUDA 12 and GPU-specific dependencies (libraft-cu12, rmm-cu12, cuda-bindings).
- Requires Python 3.11+.
License · maintenance · safety
Apache-2.0 (permissive) — Licensed under Apache-2.0 (permissive), allowing commercial and private use with minimal restrictions.
last release 2026-08-06 (8 days) · last repo commit 2026-08-14 · 1,037 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 395,975 downloads/mo, #6,976 on PyPI
Alternatives
Verify before relying
pip install pylibraft-cu12
import numpy
from pylibraft import device_ndarray
# Create GPU-resident arrays via numpy or __cuda_array_interface__ compatible objects
output = device_ndarray()- Specific CUDA compute capability requirements beyond CUDA 12 availability.
- Whether libraft-cu12 and rmm-cu12 are available for all target platforms or if installation may fail on unsupported architectures.
- Concrete example workflows demonstrating the package's primitives in isolation without external GPU libraries.
What it is and what it does
pylibraft-cu12 is the Python interface to RAFT, a C++ header-only template library of CUDA-accelerated primitives for machine learning and data mining. It exposes runtime APIs that do not require a CUDA compiler, making GPU-accelerated algorithms accessible from Python without compilation overhead.
The package provides low-level building blocks—linear algebra, matrix operations, sparse and dense computations, solvers, and statistics—designed for application developers and data source providers building high-performance GPU workflows. It integrates with the RAPIDS ecosystem via RMM (memory management) and numpy, and accepts any object supporting the __cuda_array_interface__, enabling interoperability with other GPU libraries.
Use it for
- Accelerate linear algebra operations (SVD, eigenvalue, factorization) on GPU-resident matrices.
- Implement sparse matrix operations and graph algorithms on GPU with centralized, optimized primitives.
- Develop high-performance machine learning applications that reuse vetted, maintained RAFT kernels.
- Build distributed GPU algorithms with multi-node multi-GPU infrastructure via raft-dask.
- Integrate GPU-accelerated computations into Python workflows with zero-copy interoperability.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you have a CUDA 12 GPU and need low-level accelerated primitives for algorithm development.
The active maintenance, permissive Apache-2.0 license, and interoperability with numpy and other GPU libraries make it a solid foundation for GPU-accelerated workflows. Medium install friction is acceptable for the performance and code-reuse benefits. Not suitable if you lack GPU hardware or need high-level data science tools.
Install
pylibraft-cu12 on PyPI
Before you install
Medium install friction due to CUDA 12 and GPU-specific dependencies (libraft-cu12, rmm-cu12, cuda-bindings). Requires Python 3.11+. Active maintenance with recent releases; repository shows 1037 stars and last commit on 2026-08-14.
Requires NVIDIA GPU with CUDA 12 support and CUDA Toolkit installed; Python 3.11 or later.
License in practice
Licensed under Apache-2.0 (permissive), allowing commercial and private use with minimal restrictions.
Quickstart
pip install pylibraft-cu12
import numpy
from pylibraft import device_ndarray
# Create GPU-resident arrays via numpy or __cuda_array_interface__ compatible objects
output = device_ndarray()Verify before relying
- Specific CUDA compute capability requirements beyond CUDA 12 availability.
- Whether libraft-cu12 and rmm-cu12 are available for all target platforms or if installation may fail on unsupported architectures.
- Concrete example workflows demonstrating the package's primitives in isolation without external GPU libraries.
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release >=3.11 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 4 packagescuda-bindingslibraft-cu12numpyrmm-cu12 |
| Maintenance | Actively maintained 8 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 395,975 / month, #6,976 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Intended Audience :: DevelopersProgramming Language :: PythonProgramming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14 |
Evidence: pylibraft_cu12-26.8.0-cp311-abi3-manylinux_2_24_aarch64.manylinux_2_28_aarch64.whl; pylibraft_cu12-26.8.0-cp311-abi3-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “cuda accelerated linear algebra python”
- pylibraft-cu12pylibraft-cu12 provides Python bindings to RAFT's CUDA-accelerated…
- libraft-cu12libraft-cu12 provides CUDA-accelerated primitives and algorithms for…
- nvidia-cublas-cu11Provides NVIDIA CUBLAS native runtime libraries for CUDA 11, enabling…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also libraft-cu12 · raft-dask-cu12 · pylibcudf-cu12 · libcuvs-cu12 · cuvs-cu12 · cuml-cu12 · nvidia-cusolver-cu12 · nvidia-cusparse-cu12 · libcuml-cu12 · nvidia-cusolver-cu11