speech_openai_tts

OpenAI text-to-speech exposed through the provider-neutral speech_core contracts.

Synthesis returns a cancellable AudioSource. Its PCM can be routed to playback, recording, metering, or any other bounded Audio Kit branch without giving the provider ownership of output devices.

import 'package:speech_openai_tts/speech_openai_tts.dart';

final tokenSource = StaticOpenAiTokenSource(
  OpenAiAccessToken(value: apiKey),
);

Credentials are injected through token sources, transport operations are cancellable, and HTTP/provider failures map to stable SpeechFailure values.

See the Audio Kit repository for voice orchestration examples.