cupy-cuda13x
CuPy: NumPy & SciPy for GPU
Decision gist · record as of 2026-08-14
Yes, if you have NVIDIA GPU hardware and CUDA 13.x available, and your workload involves large-scale numerical computation that would benefit from GPU acceleration. The package is production-stable, actively maintained, and has no known vulnerabilities. Install friction is moderate—you must have CUDA 13.x or use the [ctk] tag—but the API compatibility with NumPy makes adoption straightforward for existing code.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- CUDA Toolkit 13.x must be installed locally, or use [ctk] tag to bundle CUDA components from PyPI.
- Medium install friction due to compiled wheels and CUDA 13.x system requirement.
- The package is actively maintained with recent releases, and wheels are available for Python 3.10–3.14 on Linux and Windows; you must have CUDA Toolkit 13.x installed locally, or install with the [ctk] tag to bundle CUDA components.
License · maintenance · safety
MIT (permissive) — MIT license is permissive and imposes no restrictions on commercial or private use, modification, or redistribution.
last release 2026-06-01 (74 days) · last repo commit 2026-08-14 · 12,244 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 620,674 downloads/mo, #5,724 on PyPI
Alternatives
Verify before relying
pip install cupy-cuda13x[ctk]
import cupy as cp
x_gpu = cp.array([1, 2, 3])
result = cp.sum(x_gpu)- Performance gains relative to CPU NumPy for specific workload types and array sizes.
- Compatibility of CUDA 13.x with your GPU hardware and driver versions.
- Whether the [ctk] installation option fully satisfies CUDA dependencies on all platforms.
What it is and what it does
CuPy is a GPU-accelerated array computing library that mirrors NumPy and SciPy's API, allowing you to run numerical computations on NVIDIA GPUs via CUDA 13.x. It is designed for scientists and engineers who want to leverage GPU parallelism without rewriting their numerical code from scratch. The package depends on numpy and cuda-pathfinder, and is distributed as precompiled wheels for Python 3.10 through 3.14 on Linux and Windows.
You install it either by ensuring CUDA Toolkit 13.x is already on your system, or by using the [ctk] tag to have pip bundle the necessary CUDA runtime components. Once installed, you can replace NumPy arrays with CuPy arrays and call the same functions, with computation happening on the GPU instead of the CPU.
Use it for
- Accelerate large-scale matrix operations and linear algebra for machine learning or scientific simulations.
- Speed up image processing pipelines by running convolutions and transformations on GPU.
- Run iterative numerical algorithms such as optimization or PDE solvers that benefit from GPU parallelism.
- Prototype GPU-accelerated code in Python before moving to lower-level implementations.
- Process high-dimensional data arrays in applications where CPU throughput is a bottleneck.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you have NVIDIA GPU hardware and CUDA 13.x available, and your workload involves large-scale numerical computation that would benefit from GPU acceleration.
The package is production-stable, actively maintained, and has no known vulnerabilities. Install friction is moderate—you must have CUDA 13.x or use the [ctk] tag—but the API compatibility with NumPy makes adoption straightforward for existing code.
Install
cupy-cuda13x on PyPI
Before you install
Medium install friction due to compiled wheels and CUDA 13.x system requirement. The package is actively maintained with recent releases, and wheels are available for Python 3.10–3.14 on Linux and Windows; you must have CUDA Toolkit 13.x installed locally, or install with the [ctk] tag to bundle CUDA components.
CUDA Toolkit 13.x must be installed locally, or use [ctk] tag to bundle CUDA components from PyPI.
License in practice
MIT license is permissive and imposes no restrictions on commercial or private use, modification, or redistribution.
Quickstart
pip install cupy-cuda13x[ctk]
import cupy as cp
x_gpu = cp.array([1, 2, 3])
result = cp.sum(x_gpu)Verify before relying
- Performance gains relative to CPU NumPy for specific workload types and array sizes.
- Compatibility of CUDA 13.x with your GPU hardware and driver versions.
- Whether the [ctk] installation option fully satisfies CUDA dependencies on all platforms.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 2 packagesnumpycuda-pathfinder |
| Maintenance | Actively maintained 74 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 620,674 / month, #5,724 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 5 - Production/StableIntended Audience :: DevelopersIntended Audience :: Science/ResearchOperating System :: Microsoft :: WindowsOperating System :: POSIXProgramming Language :: CythonProgramming Language :: PythonProgramming Language :: Python :: 3Programming Language :: Python :: 3 :: OnlyProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: Scientific/EngineeringTopic :: Software Development |
Evidence: cupy_cuda13x-14.1.1-cp310-cp310-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp310-cp310-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp310-cp310-win_amd64.whl; cupy_cuda13x-14.1.1-cp311-cp311-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp311-cp311-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp311-cp311-win_amd64.whl; cupy_cuda13x-14.1.1-cp312-cp312-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp312-cp312-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp312-cp312-win_amd64.whl; cupy_cuda13x-14.1.1-cp313-cp313-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp313-cp313-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp313-cp313-win_amd64.whl; cupy_cuda13x-14.1.1-cp314-cp314-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp314-cp314-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp314-cp314t-manylinux2014_aarch64.whl; cupy_cuda13x-14.1.1-cp314-cp314t-manylinux2014_x86_64.whl; cupy_cuda13x-14.1.1-cp314-cp314-win_amd64.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 › “GPU array library NumPy compatible”
- cupy-cuda13xCuPy provides a NumPy and SciPy-compatible array library that runs…
- cupy-cuda12xCuPy provides a NumPy- and SciPy-compatible array library that runs…
- fast-array-utilsConverts and computes statistics across sparse, dense, GPU, and…
Give your agent the search over MCP, or paste the wish link into any chat.
More Software Development packages
Provides backported and experimental type hints for Python 3.9+, allowing use of newer typing features on older Python versions and enabling early experimentation with type system PEPs before they enter the standard library.
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.
FastAPI is a Python web framework for building REST APIs using type hints, with automatic request validation, serialization, and interactive API documentation.
Provides a way to document function parameters, class attributes, return types, and variables inline using Python's `Annotated` type hint syntax instead of traditional docstrings.
Typer builds command-line applications from Python functions using type hints, automatically generating help text, argument parsing, and shell completion.
Install it if you are building CLIs in Python.
Distlib provides low-level packaging utilities for building, distributing, and managing Python software—including metadata handling, version specifiers, wheel support, script installation, and dependency resolution.
See also cupy-cuda12x · nvidia-cusolver-cu11 · nvidia-cufft-cu12 · array-api-compat · nvidia-cufft · nvidia-cufft-cu11 · nvidia-cublas-cu11 · cutensornet-cu13 · nvidia-cudnn-cu13 · nvidia-cuda-runtime-cu12