openai_tts

A lightweight Flutter package that integrates with OpenAI's Text-to-Speech API. Supports both real-time streaming (via PCM) and downloadable audio (MP3), powered by OpenAI's latest voice models.

🎙️ Generate lifelike speech with minimal latency and stream it directly in your app.

Features

🔊 Text-to-speech using OpenAI’s tts-1 and tts-1-hd models
🎵 Supports both mp3 and real-time pcm audio streaming
🚀 Built-in streaming playback via flutter_sound
🔁 Control over playback (pause, resume, stop)
🎭 Supports multiple realistic voices

Installation

Add the following to your pubspec.yaml:

dependencies:
  openai_tts: ^<latest_version>

Then run:

flutter pub get

Getting Started

Import the package:

import 'package:openai_tts/openai_tts.dart';

Create an instance:

final tts = OpenaiTTS(apiKey: 'sk-...'); // Use your OpenAI API key

1. Download and use MP3:

final audioData = await tts.createSpeak("Hello, world!");
// Save or play using your preferred audio player

2. Stream audio in real time:

await tts.streamSpeak("Streaming this sentence in real-time.");

3. Control playback:

await tts.pausePlayer();
await tts.resumePlayer();
await tts.stopPlayer();

Configuration

Change Voice:

tts.setVoice = OpenaiTTSVoice.shimmer;

Available voices:

alloy
echo
fable
onyx
nova
shimmer
ash
ballad
coral
verse
sage

Change Model:

tts.setModel = OpenaiTTSModel.tts1hd;

Models:

tts-1
tts-1-hd

Requirements

Flutter 3.10 or newer
Valid OpenAI API key with TTS access
Internet connection

Example

Full example:

void main() async {
  final tts = OpenaiTTS(apiKey: 'sk-your-api-key');
  tts.setVoice = OpenaiTTSVoice.nova;
  tts.setModel = OpenaiTTSModel.tts1;

  await tts.streamSpeak("Welcome to OpenAI TTS with Flutter.");
}

Troubleshooting

Make sure your OpenAI API key is valid and has TTS access.
Streaming works best on physical devices with low audio latency.
Use the createSpeak method if you experience issues with real-time playback.