Get Started With the Latest NVIDIA DeepStream SDK

DeepStream SDK 9.1

Download DeepStream   Forum   Documentation  

Overview

NVIDIA DeepStream is a powerful SDK that lets you use GPU-accelerated technology to develop end-to-end, real-time streaming vision AI pipelines and agents powered by generative AI. The latest release, DeepStream 9.1, provides new skills and tools packed with innovative features to accelerate the development of your next-generation applications.

Release Highlights


Release notes
  • Access to 13 agentic skills helps fast-track vision AI development.
  • A multi-camera 3D tracking (MV3DT) skill enables accurate, end-to-end tracking across multiple cameras.
  • The AutoMagicCalib (AMC) skill automates camera calibration, saving time and reducing errors.
  • NVIDIA JetPack™ 7.2 support accelerates vision AI on the NVIDIA Jetson™ edge AI platform (Orin™, Thor™).
  • Open-source access in a unified GitHub repo simplifies adoption and maintenance.

Supported Platforms

DeepStream is available in three different flavors of containers:

  • x86
  • ARM SBSA (NVIDIA DGX Spark™)
  • Jetson (requires JetPack 7.2)

Assets
  • Starting with DeepStream 9.1 all source code files, skills, tools and installers are available on GitHub
  • Introducing ansible playbooks to provision Ubuntu hosts for x86 and ARM SBSA
  • Containers are available on NGC.

Download DeepStream SDK Containers and Installers

I Agree to the Terms of the NVIDIA DeepStream SDK Software License Agreement and to the NVIDIA AI Product License to enable download links.

DeepStream 9.1 applications are fully compatible with DeepStream 9.0. Please read the migration guide for more information.




DeepStream Skills

Connect your favorite coding agent and build production- ready applications for your vision AI use case. DeepStream Skills create Pythonic applications using DeepStream Python Service Maker.



Start Here

Note: The Python legacy bindings source code and pre-built wheels are now available on GitHub.


Get Started

Find everything you need to start developing your vision AI applications with DeepStream, including documentation, tutorials, and reference applications.

Quick Start Guide

Get step-by-step instructions for building vision AI pipelines using DeepStream and Jetson or discrete GPUs.


Get Started

Coding Agent

Generate complete DeepStream pipelines with Claude Code or Cursor using natural language prompts, reducing coding time from eight weeks to eight hours


Learn More Watch the Video

Multi-Camera Tracking

Learn how to add multi-view 3D tracking to your DeepStream pipelines.

Learn More Read the Blog

Getting Started with GitHub


Get Started

Python Service Maker

DeepStream Service Maker offers a package tailored for Python developers, providing essential bindings for pipeline APIs alongside the high-level flow APIs.


Get Started

Inference Builder

Learn how to build DeepStream apps with a declarative approach using DeepStream Inference Builder.


Get Started



Additional Resources

Technical Blogs

Webinars and GTC

DLI

Video Tutorial

Forum & FAQ

DeepStream Forums

DeepStream Forum

Tech FAQ

Technical FAQ




Ethical AI
NVIDIA platforms and application frameworks enable developers to build a wide array of AI applications. Consider potential algorithmic bias when choosing or creating the models being deployed. Also, work with the model’s developer to ensure that it meets the requirements for the relevant industry and use case; that the necessary instructions and documentation are provided to understand error rates, confidence intervals, and results; and that the model is being used under the conditions and in the manner intended.