Text-to-speech for AI agents: turn text or SSML into MP3 voiceovers with 500+ voices in 150+ locales, clone a voice from a sample, or design one from a description. Lazybird Speech API.