For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
GitHubCUDA-X
    • Home
    • Installation
      • C
      • C++
      • Python
      • Java
      • Go
      • Rust
  • Getting Started
    • Introduction
    • Integrations
    • Use-cases
  • User Guide
    • API Guide
    • Benchmarking Guide
    • Field Guide
    • References
  • Developer Guide
    • Coding Guidelines
    • Contributing
  • API Reference
    • C API Documentation
    • Cpp API Documentation
    • Python API Documentation
    • Java API Documentation
    • Rust API Documentation
    • Go API Documentation
NVIDIANVIDIA
Developer-friendly docs for your API
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2026, NVIDIA Corporation.

LogoLogocuVS
GitHubCUDA-X
On this page
  • Install Pre-Compiled Dependencies
  • Build From Source
Installation

Rust Installation

||View as Markdown|
Previous

Go

Next

Getting Started

Use this page when you need the NVIDIA cuVS Rust bindings. The Rust crate depends on the native NVIDIA cuVS C API, so libcuvs.so and libcuvs_c.so must be available when compiling and running Rust code.

All NVIDIA cuVS routine implementations live in the C++ core. The Rust bindings call into the native C and C++ libraries, so install both libcuvs_c and libcuvs.

Install Pre-Compiled Dependencies

Install the native libraries first:

$# CUDA 13
$conda install -c rapidsai -c conda-forge libcuvs cuda-version=13.2
$
$# CUDA 12
$conda install -c rapidsai -c conda-forge libcuvs cuda-version=12.9

When building examples or working on the Rust bindings from source, use the centralized build environment guidance. A Rust-specific environment YAML file is also available in conda/environments when you want a narrower development environment.

Configure LIBCLANG_PATH so Rust bindgen can find libclang:

$LIBCLANG_PATH=$(dirname "$(find "$CONDA_PREFIX" -name libclang.so | head -n 1)")
$export LIBCLANG_PATH

Add NVIDIA cuVS to your Rust project:

1[dependencies]
2cuvs = ">=24.6.0"

Then build or run your project with Cargo:

$cargo build

Build From Source

Before building from source, review the shared C++ source-build prerequisites, including the recommended conda environment setup for build dependencies.

Build the native libraries first if needed:

$./build.sh libcuvs

Then build the Rust bindings:

$./build.sh rust