OpenStudio

What OpenStudio is

OpenStudio is a self-hosted virtual studio for live call-in broadcasts, multi-host podcasts, and community radio. It runs on the web, with peer-to-peer WebRTC mesh audio and mix-minus monitoring for around 15 participants. Operators can send audio through Icecast to unlimited listeners using their own domain, while self-hosted broadcasts have unlimited airtime.

The browser records individual voice tracks alongside a program mix. Exports include WAV and WebM, with MP3 available when ffmpeg is installed. Optional post-production tools add on-device transcription with whisper.cpp, audio cleaning, and LLM-generated show notes. Show notes can connect to OpenAI-compatible endpoints such as LM Studio, Ollama, OpenAI, Together AI, and Groq; Anthropic requires a compatibility shim.

The project describes token-based rooms and invites, role-based access, connection and message limits, rate limits, CORS controls, and security headers. It says audio stays on infrastructure controlled by the operator and characterizes the project as having zero tracking. The quick start uses Git, npm, and a local browser address. The live demo expires broadcasts after 15 minutes; self-hosted broadcasts do not have that airtime limit.

Who it suits

OpenStudio suits creators, community radio hosts, and podcast collectives that want to run live audio sessions on infrastructure they control. It is also a fit for operators who need separate participant tracks and streaming to an audience through Icecast.

Limits

The live demo automatically ends broadcasts after 15 minutes, while self-hosted broadcasts have unlimited airtime. Mesh audio supports around 15 participants. Optional transcription and audio cleaning require ffmpeg, ffprobe, cmake, and about 1.5 GB of free disk space for the default Whisper model.

Getting started

1. Clone the project from its GitHub website.
2. Install dependencies with npm install.
3. Start the app with npm start.
4. Open localhost:6736 in a browser.
5. For optional transcription and cleaning, install ffmpeg, ffprobe, and cmake, and allow about 1.5 GB for the default Whisper model.

Verdict

Choose OpenStudio if you want a self-hosted studio for live audio, with browser recording, individual tracks, and Icecast streaming. Look elsewhere if you need more than around 15 participants or a host interface that creates invite links.