Agora's Real-Time Speech to Text is an extension that provides accurate live transcription and subtitling services. It enables the conversion of audio to text for active or selected hosts in real time, with the text distributed as live captions to all participants. This service integrates seamlessly with Large Language Models (LLMs) for further processing, without impacting real-time communication performance. It supports transcribing and labeling simultaneous speakers, even with up to three concurrent speakers, and offers captioning for cloud recordings. With multi-language support for all major languages and dialects, and enterprise-grade security and compliance, Agora's Real-Time Speech to Text enhances accessibility, engagement, and analysis across various applications.

Features:

• Cloud-based live transcription: Cloud-based transcription converts audio to text for active or selected hosts in real time. Text can be distributed as live captions to all participants in the channel.
• LLM integration: Integrate speech to text with LLMs for further processing, without impacting RTC performance. Upload transcription text as .vtt files to LLMs like GPT to generate summaries, notes, and more.
• Transcribing and labeling simultaneous speakers: Easily label who said what—even with up to 3 simultaneous speakers. Separate transcription for each host ensures accuracy and allows you to choose to transcribe for one specific host.
• Captioning for cloud recordings: Transcribe audio to text on video or audio recordings to enable closed captions (CC) on playback or review important discussion items in the transcript.
• Multi-language support: Real-time transcription supports all major languages and dialects, and each channel can support audio-to-text transcription for up to two languages simultaneously.
• Enterprise-grade security and compliance: Agora is ISO and SOC 2 certified and meets compliance standards for regional privacy laws and industry regulations, including GDPR, CCPA, and HIPAA. Live captions and transcription can be encrypted in the same way as encrypted RTC audio or video.

Reduce cost and increase efficiency More efficient and cost-effective than traditional client-side live transcription, Agora’s solution by uses advanced technology to remove silence, reduce Word Error Rate (WER), and distribute live captions to all participants in a channel. Get the most accurate results at scale Cutting-edge AI ensures the highest accuracy even with overlapping speech, regional accents, and poor network conditions. Scale from one-to-one meetings to up to millions of participants with the same accuracy. Integrate with ease Agora’s Real-Time Speech to Text is highly integrated with Agora’s network (SD-RTN™), providing global user transcription and real-time text distribution even in poor network environments.