Skip to main content

Voxbento: Real-Time Interpretation for Live Events

Voxbento lets you run professional simultaneous interpretation for any live event, entirely in the browser. Interpreters monitor the floor session through a built-in Jitsi console and broadcast translated audio directly to attendees via WebRTC — with sub-second latency and seamless handoff between interpreter shifts.

Quick Start​

Set up your first event, booth, and invite interpreters in minutes.

How It Works​

Understand the interpretation flow from floor audio to listener playback.

Admin Guide​

Create events, manage rooms, assign roles, and generate invite tokens.

API Reference​

Integrate Voxbento into your event platform using the REST and WebSocket APIs.

How Voxbento Works​

Create an event and booths​

Log in to the admin panel and set up your event with rooms and language booths. Each booth represents one interpretation channel (e.g., English, French, Spanish).

Invite interpreters and coordinators​

Generate invite tokens for each booth and share the links. Participants open their unique link — no account needed — and join the booth directly in their browser.

Go live​

The active interpreter clicks Go Live. Audio streams instantly via WebRTC/WHIP to MediaMTX and is served to listeners with sub-second latency.

Listeners tune in​

Attendees visit the listener page for your event, pick their language, and hear the interpretation in real time — with optional live captions powered by AI transcription.

Key Features​

Sub-second Latency​

WebRTC/WHEP delivery keeps audio delay under one second for every listener.

Seamless Handoff​

Coordinators switch active interpreters without dropping listeners or interrupting playback.

Live Captions & AI Translation​

Automatic transcription and real-time translation powered by LLMs (OpenAI, Anthropic, Groq, Deepgram, Gemini).

Floor Audio Support​

Capture the main stage audio seamlessly with the built-in Floor Bot to provide captions and translations for the original speaker.

No Install Required​

Everything runs in the browser. Interpreters and listeners need no plugins or apps.

Multi-language​

Run unlimited language booths simultaneously for a single event.

Role-based Access​

Fine-grained roles — coordinator, interpreter, listener — scoped per event or per booth.