• 200+ pretrained models from Open Model Zoo for the OpenVINO toolkit for a wide variety of use cases.
  • Seamless Hugging Face integration for an expansive range of Generative AI (GenAI) models and large language models (LLM).
  • Option to import custom models from PyTorch*, TensorFlow*, and ONNX* (Open Neural Network Exchange)
  • Built-in OpenVINO toolkit AI inference runtime optimizations and benchmarking
  • Performance data for different topologies and layers
 
  • Standardized development interfaces: JupyterLab and Microsoft Visual Studio* code IDEs for elevated coding experience.
  • Ready-to-use reference implementations: Preconfigured, use-case-specific applications with the complete stack of reusable software.
  • OpenVINO toolkit samples and notebooks: Computer vision, generative AI, and LLM use cases.
  • Diverse component integration: Importing source code and native applications, Docker* containers, and Helm* charts directly through any popular repositories.

 

 
  • Benchmarking on hybrid computing architectures including CPUs, GPUs, and AI accelerators.
  • Single and multi-node benchmarking to analyze workloads in isolation or at scale.
  • Application and system telemetry for data-driven decision-making.