Hub
Voice + audio hub
Everything you need to build voice experiences with Meridian. From raw speech primitives to full conversational agents.
Text-to-Speech
Convert text into natural speech with 40+ voices across 12 languages. Stream or download WAV/MP3 output.
Read docs→
Speech-to-Text
Transcribe audio files or live streams with word-level timestamps and speaker diarization.
Read docs→
Voice Agent Recipe
End-to-end guide for building a voice agent: STT → LLM → TTS pipeline with WebSocket streaming.
Read docs→
Recipe Voice Form
Collect structured data over a phone call. Define a JSON schema and let the agent ask questions naturally.
Read docs→
Quick start
New to voice? Start with Text-to-Speech to generate your first audio file, then follow the Voice Agent Recipe to wire up a full conversational pipeline.