Do you offer voice-to-voice or text-to-speech (TTS)?

Yes. AssemblyAI Streaming TTS turns text into speech with the assemblyai-tts-1 model, built for conversational voice agents. You can stream audio over a WebSocket (one session per conversation) or over HTTP, and OpenAI-compatible endpoints let you keep code written for the OpenAI speech API. See the Streaming TTS quickstart to get started. Streaming TTS is billed at $15 per 1M input characters. See pricing for details. For a full voice-to-voice agent, pair Streaming TTS with Universal-3.6 Pro Realtime streaming speech-to-text and an LLM, for example through LLM Gateway. See Best practices for building voice agents, and use the LiveKit or Pipecat integration to run the whole pipeline in your framework.