---
id: custatevec-cu13
version: "1.14.0"
license: NVIDIA Proprietary Software
license_treatment: unclear
maintenance: active
---

# custatevec-cu13 — cuStateVec - a component of NVIDIA cuQuantum SDK

License: unclear · Maintenance: active · Downloads: 92.8K/mo

## What it is and what it does

cuStateVec is NVIDIA's high-performance library for state vector quantum simulation operations, part of the cuQuantum SDK. It provides GPU-accelerated computation on NVIDIA CUDA hardware, enabling researchers and developers to build quantum simulators that leverage GPU parallelism for faster state vector manipulations and measurements.

The package is distributed as platform-specific wheels for Linux (x86_64 and aarch64) and requires CUDA 13 to be installed on the system. It has no Python runtime dependencies, meaning once installed, it provides direct access to the underlying C library. The documentation notes that for Python-level APIs, users should install the cuQuantum Python wrapper instead, suggesting this package is primarily a lower-level component.

Use it for:

- Building GPU-accelerated quantum circuit simulators that need fast state vector operations
- Prototyping quantum algorithms on NVIDIA GPUs without writing CUDA C code directly
- Integrating state vector quantum simulation into larger scientific computing pipelines on GPU clusters
- Research into quantum algorithm performance on modern GPU hardware

## Worth the install?

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

cuStateVec provides GPU-accelerated state vector operations for quantum simulators, optimized for NVIDIA CUDA 13 hardware.

Yes, if you are developing quantum simulators on NVIDIA CUDA 13 hardware and need GPU acceleration. The package is actively maintained, has no known vulnerabilities, and fills a specialized role in the quantum computing ecosystem. However, verify the NVIDIA Proprietary Software license terms for your use case first, and confirm your system has CUDA 13 and a compatible GPU before installing.

## Install

pip install custatevec-cu13
uv add custatevec-cu13
poetry add custatevec-cu13

## Installing custatevec-cu13

Before you install: The package is actively maintained with recent releases and carries medium install friction due to platform-specific wheels (x86_64 and aarch64 Linux only). No runtime dependencies simplify deployment once the wheel is available for your architecture.

License in practice: Licensed under NVIDIA Proprietary Software with no SPDX identifier; the exact terms and redistribution rights are unclear from the metadata alone and should be verified against NVIDIA's licensing documentation before production use.

Quickstart:

pip install custatevec-cu13

import custatevec
# Access cuStateVec C API or use via cuQuantum Python wrapper

Requires NVIDIA CUDA 13 and compatible GPU; wheels are only available for Linux x86_64 and aarch64 architectures.

Verify before relying:

- Exact licensing terms and permitted use cases under NVIDIA Proprietary Software license
- Whether Python API is directly exposed or requires cuQuantum Python wrapper
- Minimum Python version requirement (not specified in metadata)

## Package facts

- License: NVIDIA Proprietary Software (unclear)
- Python support: unspecified
- Install friction: medium
- Maintenance: active
- Downloads: 92.8K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

quantum simulator gpu acceleration, state vector cuda library, nvidia custatevec, quantum computing high performance, cuda quantum state operations, gpu-acceleration, quantum-computing, cuda

[View on SkillFed](https://skillfed.io/packages/custatevec-cu13) · [View on PyPI](https://pypi.org/project/custatevec-cu13/)