nvidia-cuda-runtime-cu11

What it is and what it does

This package bundles NVIDIA's CUDA 11 runtime native libraries as a Python wheel, allowing GPU-accelerated applications to access the low-level CUDA runtime without requiring a separate CUDA toolkit installation. It is typically installed as a dependency of higher-level frameworks that need GPU support.

The package provides platform-specific binaries for Linux (x86_64 and aarch64) and Windows, selected automatically during installation. However, the package has been abandoned with no active maintenance since 2022-10-03, meaning it will not receive updates for newer CUDA versions, Python releases, or security patches. Users relying on this package should be aware that it is frozen at version 11.8.89 and may become incompatible with future system configurations.

Use it for:

• Installing as a transitive dependency of deep learning frameworks that require CUDA runtime support.
• Setting up GPU-accelerated compute environments on Linux or Windows without manually installing the full CUDA toolkit.
• Containerized deployments where CUDA runtime libraries need to be bundled with Python packages.
• Legacy projects locked to CUDA 11 that cannot upgrade to newer CUDA versions.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides CUDA 11 runtime native libraries for GPU-accelerated computing on Linux and Windows systems.

No—unless you are maintaining legacy code locked to CUDA 11. The package is abandoned with no maintenance since 2022-10-03, no repository activity, and unclear proprietary licensing. For new projects, use actively maintained CUDA runtime packages or install CUDA directly from NVIDIA. For existing deployments, verify that your framework does not already provide or recommend a replacement.

Install

nvidia-cuda-runtime-cu11 on PyPI

pip install nvidia-cuda-runtime-cu11

uv add nvidia-cuda-runtime-cu11

poetry add nvidia-cuda-runtime-cu11

Installing nvidia-cuda-runtime-cu11

Before you install

Medium install friction due to platform-specific wheel distributions (x86_64 and aarch64 Linux, Windows). Package is abandoned with no active maintenance since its latest release on 2022-10-03.

License in practice

Licensed under NVIDIA Proprietary Software with unclear treatment—not an SPDX standard license. Users should verify licensing terms directly with NVIDIA before deploying in production or commercial contexts.

Quickstart

pip install nvidia-cuda-runtime-cu11==11.8.89

# Import in code that uses CUDA-dependent libraries
import nvidia.cuda_runtime

Requires a compatible NVIDIA GPU and CUDA-capable system; Windows and Linux (x86_64/aarch64) only. Package is abandoned and may not receive updates for newer CUDA or Python versions.

Verify before relying

• Whether this package receives security updates or maintenance from NVIDIA.
• Compatibility with Python versions beyond 3.11 and future CUDA toolkit releases.
• Whether the proprietary license permits redistribution in commercial or closed-source applications.

Package facts

Evidence: nvidia_cuda_runtime_cu11-11.8.89-py3-none-manylinux1_x86_64.whl; nvidia_cuda_runtime_cu11-11.8.89-py3-none-manylinux2014_aarch64.whl; nvidia_cuda_runtime_cu11-11.8.89-py3-none-manylinux2014_x86_64.whl; nvidia_cuda_runtime_cu11-11.8.89-py3-none-win_amd64.whl

Keywords: cuda, nvidia, runtime, machine learning, deep learning

Tags

More Software Development packages

Further reading