AgentSea.Voice.TTS behaviour (agentsea_voice v0.1.0)

Copy Markdown

Text-to-speech behaviour. Adapters: remote (OpenAI/ElevenLabs over Req) or, in future, local (Piper / Bumblebee).

Summary

Types

result()

@type result() :: %{audio: binary(), format: String.t()}

Callbacks

synthesize(text, opts)

@callback synthesize(text :: String.t(), opts :: keyword()) ::
  {:ok, result()} | {:error, term()}