Agora's Real-Time Translation is an extension that provides accurate live translation with ultra-low latency. It enables seamless global communication by translating spoken content into up to five target languages, supporting 30+ languages. The solution uses advanced speech recognition (ASR) to accurately capture spoken language and convert it to text. Translated live captions are continuously delivered to all participants, and video text track (VTT) files can be stored for future reference. It ensures seamless translation for real-time communication with end-to-start latency of less than 1 second and an average end-to-end latency of under 3 seconds. The translated text can be processed using custom large language models (LLMs) or integrated with additional AI services to enhance capabilities and streamline workflows. Agora’s Real-Time Translation helps remove language barriers, expand your audience, and enable global connections.

Features:

• Live translation: Live speech-to-text translation to keep the conversation flowing seamlessly in real-time communication or live streaming.
• Multi-language translation: Manage multilingual interactions with speech translation of up to two source languages into five target languages with support for 30+ languages.
• High accuracy: Advanced Speech Recognition (ASR) captures spoken language and converts it to text accurately using sophisticated speech recognition technologies.
• Translated captions: Easily readable translated live captions are continuously delivered to all participants. Video text track (VTT) files can be stored in the cloud for future reference, AI analysis, or compliance.
• Ultra-low latency translation: Ensure seamless translation for real-time communication with end-to-start latency of less than 1 second and an average end-to-end latency of under 3 seconds.
• LLM integration: Process translated text using custom large language models (LLMs) or integrate with additional AI services to enhance capabilities and streamline workflows.