TechPrufer
Get started
NVIDIA
NVIDIA logoNVIDIA · Feature trail

Agentic Video Skills for Jetson

1 post · Jun 2026 – Aug 2026

What this trail covers

This feature thread tracks the evolution of agentic video skills for NVIDIA Jetson platforms, enabling programmable, device-aware video workflows. Initial efforts focused on providing foundational agentic skills above the Video Codec SDK and PyNvVideoCodec, connecting developer intent to live device discovery, supported configurations, reproducible execution, and performance verification. JetPack 7.2.1 introduces foundational video workflows through the unified jetson-videosdk skill, allowing developers to automate tasks like Linux customization, memory optimization, and model benchmarking. JetPack 7.2 also introduces Multi-Instance GPU (MIG) support on Jetson Thor for isolated, predictable multiworkload execution, critical for real-time AI applications in robotics and industrial automation. Official Yocto Project support and Super Mode for Jetson AGX Orin 32 GB further enhance system efficiency and AI performance.

All NVIDIA trails
  1. NVIDIA JetPack 7.2.1 Adds Agentic Video Skills and T3000 Emulation | NVIDIA Technical BlogJetPack 7.2.1 introduces foundational agentic video skills for Jetson platforms, building upon the Video Codec SDK and PyNvVideoCodec. These skills enable programmable, device-aware video workflows by connecting developer intent to live device discovery, supported configurations, reproducible execution, and performance verification. The unified jetson-videosdk skill allows developers to discover and report capabilities, generate encoder recipes, benchmark performance and quality, and validate codec workflows. PyNvVideoCodec 2.2 support on Jetson provides Python-based, hardware-accelerated video encoding and decoding with DLPack integration and features like ThreadedDecoder for improved pipeline efficiency. The release also enables emulation of Jetson T3000 performance on Jetson T5000 using the Jetson Thor AGX Developer Kit for rapid development.
  2. Deploy Agentic-Ready AI at the Edge with Memory Efficiency in NVIDIA JetPack 7.2 | NVIDIA Technical BlogJetPack 7.2 introduces one-command deployment of NVIDIA NemoClaw, preconfigured dependencies for agentic AI applications on Jetson. It adds Jetson agent skills (device-side and BSP-side) to automate development tasks like Linux customization, memory optimization, and model benchmarking. JetPack 7.2 also brings Multi-Instance GPU (MIG) support to Jetson Thor, enabling partitioning into two isolated GPU instances for mixed-criticality workloads, and official Yocto Project support for custom Linux distributions.