openvinotoolkit/openvino
OPTIMIZATION AND DEPLOYMENT OF AI MODELS
Toolkit for optimizing and deploying deep learning models. Supports conversion from PyTorch, TensorFlow, ONNX and inference on CPU, GPU, and Intel NPU.
// KEY FEATURES
- Converts models from PyTorch, TensorFlow, ONNX and other frameworks
- Accelerates inference on CPU (x86/ARM), Intel GPU and NPU
- Supports LLMs, diffusion, computer vision and speech
- Provides Python, C++, NodeJS and GenAI APIs
// DEEP DIVE: HOW IT WORKS
OpenVINO converts models from PyTorch, TensorFlow, ONNX, PaddlePaddle, JAX/Flax to IR format and compiles for CPU (x86/ARM), Intel GPU or NPU. Applies quantization, layer fusion and kernel selection to accelerate inference.
APIs: Python, C++, C, NodeJS, GenAI. Key difference — native Intel NPU support and Optimum Intel for Hugging Face (transformers, diffusers without frameworks). Notebooks: LLM chatbots, YOLOv11, Stable Diffusion, Whisper, LLaVa.
OS: Linux, Windows, macOS. GPU/NPU require Intel drivers (Level Zero, OpenCL). NPU only on Core Ultra (Meteor Lake+). Quantization may need calibration data. GenAI in separate repo openvinotoolkit/openvino.genai.
// USE CASES
Best for: For ML engineers deploying models on Intel hardware
- Convert PyTorch model and run inference on CPU in minutes
- Deploy LLM chatbot on local NPU without cloud
- Accelerate YOLOv11 or Stable Diffusion on Intel integrated GPU
// QUICK START
pip install -U openvino
import openvino as ov
print(ov.__version__)Commands are taken from the repository README — check the current version on GitHub before running.
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram