Details
NVIDIA Broadcast runs AI models locally on RTX GPU Tensor Cores to process audio and video feeds in real time. It creates virtual microphone and camera devices that any broadcasting or conferencing app can select as inputs. The app supports noise removal, room echo removal, background blur or replacement, Eye Contact (which repositions the speaker's gaze to appear camera-facing), Auto Frame, and video noise removal for low-light conditions. As of a January 2025 update, two additional beta effects are available: Studio Voice, which synthesizes enhanced microphone audio quality, and Virtual Key Light, which digitally relights the speaker's face; both require a GeForce RTX 4080 or RTX 50 Series GPU. An enterprise installer version is also available for IT deployment at scale.
Have evidence about NVIDIA's AI practices? Submit a report.
Report a Sighting →