Video and Audio Solutions

NVIDIA’s AI-powered SDKs for building GPU-accelerated multimedia processing applications provide real-time solutions for enhancing, modifying, and optimizing audio and video quality. Packaged into two core live streaming and video conferencing products—NVIDIA Broadcast Engine and NVIDIA Maxine™—these SDKS enable developers to improve live streaming, content creation, and video conferencing applications and platforms.

body tracking

Face and Body Tracking

The AR SDK allows developers to enable real-time 3D tracking of a person’s face and body using a standard web camera, creating fun and engaging augmented reality (AR) effects.

Get Started with AR SDK

Audio Enhancements

With the Audio Effects SDK, developers can utilize AI to remove distracting background noise, isolating human speech in audio from incoming and/or outgoing audio feeds.

Get Started with Audio Effects SDK
super res

Super Resolution and Virtual Backgrounds

The Video Effects SDK allows developers to utilize AI to upscale video resolution in real time and provide virtual backgrounds, optimize the viewing experience.

Get Started with Video Effects SDK
video analytics

Video and Image Analytics

The DeepStream SDK delivers a complete streaming analytics toolkit for AI-based multi-sensor processing and video and image understanding.

Get Started with Deepstream SDK
Conversational AI

Conversational AI

The Riva SDK is an application framework for multimodal conversational AI services that delivers real-time performance on GPUs.

Get Started with Riva SDK
facial animation

Facial Animation from Voice

The Audio2Face SDK is an AI-powered application that generates expressive facial animation from just an audio source.

Get Started with Audio2face SDK

Live Streaming for Windows Machines

NVIDIA Broadcast Engine is a suite of SDKs that provides AI-powered features for broadcasting, content creation, and video conferencing to developers utilizing AI on local Windows machines.

Learn More

Platform for Cloud-Based Video Conferencing

NVIDIA Maxine is a fully accelerated platform SDK for building and deploying AI-powered video conferencing features using state-of-the-art models in the cloud.

Learn More

Display and Output Technologies

View a collection of products for creating optimized GPU-accelerated applications for any type of display. These products enable developers to improve the delivery and synchronization of their visual applications.

Learn More

Core Video Technologies

Explore the hardware-accelerated technology that serves as the foundation for all video solutions. Taking advantage of the latest NVIDIA architecture, this technology frees up the GPU and CPU for other operations, providing more computational power for developers to create their applications.

Video Encode and Decode

The Video Codec SDK is a comprehensive set of APIs, including high-performance tools, samples, and documentation for hardware-accelerated video encode and decode on Windows and Linux.

Learn More

Graphics and Compute

The Vulkan API is a new-generation graphics and compute open-standard API that provides high-efficiency, cross-platform access to modern GPUs.

Learn More

Accurate Flow Vectors

The Optical Flow SDK uses sophisticated algorithms to yield highly accurate flow vectors, which are robust to frame-to-frame intensity variations and can track true object motion.

Learn More

High-Performance Data Transfer

The Rivermax SDK enables direct data transfers to and from the GPU, delivering best-in-class throughput and latency with minimal CPU utilization for streaming workloads.

Learn More

Accelerated Game Asset Loading and Decompression

NVIDIA RTX™ IO enables rapid GPU-based loading and game asset decompression, accelerating IO performance by up to 100X compared to hard drives and traditional storage APIs.

Learn More

Low-Latency IO and GPU/CPU Synchronization

GPUDirect® for Video allows third-party hardware to efficiently communicate with NVIDIA GPUs and minimize historical latency issues by synchronizing IO devices with the GPU and CPU.

Learn More

Direct Transfer Between GPU and Storage

GPUDirect Storage enables a direct data path for direct memory access (DMA) transfers between GPU memory and storage, which avoids a bounce buffer through the CPU.

Learn More

Neural Network Trainer

The Data Loading Library reduces latency and training time by overlapping training and pre-processing when decoding and augmenting images, videos, and speech.

Learn More

Accelerated Video and Image Processing

The Performance Primitives library provides GPU-accelerated image, video, and signal processing functions that perform up to 30X faster than CPU-only implementations.

Learn More

Low-Latency Image Decoding

The nvJPEG libraries provide low-latency decoding, encoding, and transcoding for common JPEG formats used in computer vision applications, such as image classification/segmentation and object detection.

Learn More



NVIDIA On-Demand

Explore the extensive catalog of sessions, podcasts, demos, research posters and more.

Learn More


Engage with NVIDIA directly and discuss the latest technology with a global developer community.

Visit the NVIDIA Developer Forums

NVIDIA Deep Learning Institute

The NVIDIA Deep Learning Institute (DLI) offers hands-on training in AI, accelerated computing, and accelerated data science.

Learn More

NVIDIA Developer Program

Join the NVIDIA Developer Program for access to a wide variety of tools and training to accelerate AI, HPC, and advanced graphics applications.

Learn More

Video Codec SDK GPU Support Matrix

Support for various video formats across GPUs

Learn More