Open Live Studio browser interface — multiviewer and vision mixer running in a web browser, no dedicated hardware control surface — Open Live Studio — the browser-based production UI with multiviewer and vision mixer, running on commodity GPU hardware.

The challenge

To understand why that mattered, look at where live production has stayed stuck. Vision mixers and the surrounding production chain have traditionally meant dedicated hardware — six-figure capital costs, proprietary panels and matrices, and trained operators tied to a physical control surface. For most productions that complexity is overkill, but the lack of a credible software alternative kept teams locked in.

Open Live changes that. It brings the full live-production chain — mixing, audio, multiviewer and rendering — onto commodity GPU hardware you can run in the cloud or on-premises, under an MIT licence, with no proprietary hardware dependency.

What Eyevinn built

Open Live is a cloud-native, GStreamer-based live broadcast production platform. The entire production chain runs in software on commodity GPU hardware, controlled from the browser and deployed Kubernetes-native.

Vision mixer — multi-source live switching with a browser-based UI, no dedicated control surface required.
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.
Audio engine — up to 32 channels, a SIMD Rust LSP port, EBU R128 loudness compliance, and AES67/Ravenna for professional audio-over-IP.
Multiviewer — monitor every source at a glance.
WebRTC sub-500 ms latency — low-latency contribution and monitoring over any IP network.
SRT input and output and EFP for up to 16 sources, for resilient contribution feeds.
Cloud-native and Kubernetes-native — scale on commodity hardware, deploy in minutes, no broadcast-specific appliances.
MIT-licensed open source — no licence cost, no vendor lock-in.

Open Live signal path: SRT, EFP and WebRTC WHIP ingest feeding the Strom GStreamer GPU rendering pipeline, then WebRTC WHEP output to browser viewers — Open Live signal path: SRT/EFP/WebRTC WHIP ingest → Strom GStreamer GPU pipeline → WebRTC WHEP to browser viewers.

How to get Open Live — three options

1. Managed on Open Source Cloud (hosted)

Run Open Live as a managed service on Eyevinn Open Source Cloud — a production-ready deployment in minutes, no infrastructure to maintain. Plans start at €15/month (Personal) and €69/month (Professional, excl. VAT).

Try it live →

2. Self-hosted (open source)

Open Live is fully open source under the MIT license (github.com/Eyevinn/open-live) with zero vendor lock-in. Run it on your own infrastructure, another cloud, or on-premises at no licence cost.

3. Dedicated managed instance via Eyevinn

Need a dedicated deployment, deeper integrations or a tailored setup? Our team can run Open Live as a dedicated managed instance and build the integrations you need as a paid engagement on top of the open-source core. We scope it together and quote based on your requirements. Contact sales@eyevinn.se.

Talk to the team →

Read the engineering story

We wrote up how Open Live was built — the GPU rendering pipeline, the audio engine, and the path to a software broadcast mixer on commodity hardware.

Read the engineering story on Medium →

The Open Live source is on GitHub: Eyevinn/open-live, Eyevinn/strom (engine), and Eyevinn/open-live-studio (browser UI).

Talk to the team behind Open Live

Want to run Open Live on a real production, connect it to your existing gear, or have us deploy and tailor a dedicated managed instance? Try it live yourself, or tell us about your setup and we will recommend the right option.

Try it live → Talk to us about your setup →