The broadcast switcher has been one of the last pieces of production infrastructure that resisted software-defined approaches. Cameras, graphics, playout, and contribution have all moved to software and the cloud — but the vision mixer at the heart of a live gallery has largely stayed locked to proprietary hardware.

Open Live changes that. It is a cloud-native, software-defined live broadcast production platform — vision mixer, audio engine, multiviewer, and GPU rendering pipeline — released under the MIT licence and runnable on commodity GPU hardware.

What Open Live is

Open Live is a complete live production stack you can run in the cloud or on-premises, with no licence cost and no vendor lock-in:

  • Vision mixer — software-defined switching at the centre of the gallery.
  • GPU rendering pipeline — zero-copy CUDA-GL interop and hardware encoding via NVENC, with NVIDIA, Intel Quick Sync Video (QSV), and AMD VA-API support, so it runs across the GPU hardware you already have.
  • Audio engine — up to 32 channels, EBU R128 loudness compliance, and AES67/Ravenna for professional audio-over-IP.
  • Multiviewer — monitor all your sources at a glance.
  • WebRTC sub-500 ms latency — low-latency contribution and monitoring, with SRT input and output and EFP for up to 16 sources for resilient contribution feeds.
  • GStreamer-based and Kubernetes-native — built on proven open media plumbing and designed to scale on standard cloud infrastructure.

Three ways to use it

  1. Managed on Open Source Cloud — run Open Live as a managed service on Eyevinn Open Source Cloud, with nothing to install.
  2. Self-hosted open source — the platform is MIT-licensed, so you can self-host it on your own commodity GPU hardware at no licence cost.
  3. Dedicated managed instance — for a managed, dedicated deployment, contact us at sales@eyevinn.se.

Try it and read more

The source is open on GitHub: the Open Live API, the Strom engine, and the Open Live Studio browser UI.


Interested in deploying Open Live for your live production? Contact us to discuss a managed or dedicated instance.