#9696th of 168 in ML frameworks and libraries
cv-cuda
GPU-accelerated image processing operators for vision pipelines
- Stars
- 2.7k
- License
- custom license
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- C++
2,652 on 21 Apr 20262,734 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
CV-CUDA is an NVIDIA library of CUDA-based computer vision operators, such as resize, for pre- and post-processing images and video in AI pipelines. It has C/C++ and Python APIs and works with tools like nvImageCodec, so decoded data can stay on the GPU. Wheels are published on PyPI for CUDA 12 and CUDA 13.
Who it is for: Engineers speeding up image and video pre/post-processing on NVIDIA GPUs
Strengths
- Wheels for CUDA 12 and 13 on x86_64 and aarch64 SBSA, Python 3.10 to 3.14
- C/C++ and Python APIs; tensors can come from GPU decoders like nvImageCodec
- Jetson Orin and Thor builds are available alongside server GPUs
- Compatibility table lists driver, compute capability and compiler per build
Weaknesses
- Requires an NVIDIA GPU with compute capability 7.5 or higher
- No native Windows support, only WSL2
- Only one CUDA major version of the packages can be installed at a time
- Not accepting external contributions; OSD text rendering has known issues on Jetson
What it needs
- GPU required
- Docker
- Needs CUDA 12.2+ or 13.x, NVIDIA driver r525+ (r580+ for CUDA 13), CMake 3.24.3+ and gcc 10+ to build from source
Quick answers
Is cv-cuda open source?
Its code is public on GitHub under its own license terms; read them before you use it at work.
Can you run cv-cuda with Docker?
Yes. Its repository has a Docker setup: a Dockerfile, a Compose file or a published image.
Does cv-cuda need a GPU?
Yes, it needs a GPU.
Also in ML frameworks and libraries
See all 168| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |