Real-Time Translation Guide

Before You Proceed: Managed Translation Is Available

MediaSFU provides managed translation configurations for users with valid subscriptions. You can enable this inLiteDashboard Settings. When you use managed configs, MediaSFU bills usage directly instead of using your own provider accounts. Eligible plans include monthly translation credits, and you can top up whenever you need additional capacity.

Overview

Real-Time Translation for Global Communication

TranslationConfig enables seamless multilingual communication in video conferences by combining Speech-to-Text (STT), Language Models (LLM), and Text-to-Speech (TTS) into a unified translation pipeline.

Speech Recognition

Convert spoken audio to text with high accuracy using leading STT providers

AI Translation

Leverage powerful LLMs to translate text while preserving context and nuance

Voice Synthesis

Generate natural-sounding speech in the target language with customizable voices

20+ Languages

Support for major world languages with extensible language configuration

Prerequisites:

You can complete setup fully in Dashboard usingAI CredentialsandTranslation Configs(API is optional)
Your subscription must support translation features
Select STT, LLM, and TTS providers in AI Credentials, then create your translation pipeline in Translation Configs
If you prefer API workflows, useSandbox AI CredentialsthenSandbox Translation Config

Overview

Real-Time Translation for Global Communication

Speech Recognition

AI Translation

Voice Synthesis

20+ Languages

Quick Start

Architecture

Configuration

Supported Languages

Provider Types

Pipeline Configuration

Voice Overrides

API Examples

Room Integration

Best Practices

Troubleshooting