echonova_ai 0.0.1
echonova_ai: ^0.0.1 copied to clipboard
Official Dart and Flutter SDK for the EchoNova speech API: Text to Speech, Speech to Text and Sound Cleaning, over REST and WebSocket streaming.
EchoNova examples #
Text to Speech and Speech to Text from server-side Dart. In a Flutter app, use EchoNova.proxy with your own backend instead (proxy_mode.dart): never ship an API key in an app.
// Text to Speech and Speech to Text with the EchoNova SDK, from server-side
// Dart. In a Flutter app, use EchoNova.proxy with your own backend instead:
// never ship an API key in an app.
//
// Run:
//
// ECHONOVA_API_KEY=... dart run example/echonova_example.dart
import 'dart:io';
import 'package:echonova_ai/echonova_ai.dart';
Future<void> main() async {
final client = EchoNova(); // reads ECHONOVA_API_KEY
try {
// Voices change rarely: list them once and cache them.
final voices = await client.textToSpeech.voices.list();
final voice = voices.firstWhere(
(v) => v.language == 'ar',
orElse: () => voices.first,
);
stdout.writeln('voice: ${voice.displayName ?? voice.name} (${voice.id})');
// Synthesize and download the WAV. The link expires: download it soon.
final speech = await client.textToSpeech.generate(
'مرحبا بكم في إيكونوفا',
voiceId: voice.id,
);
final audio = await client.download(speech.audioUrl!);
await File('speech.wav').writeAsBytes(audio);
stdout.writeln('speech.wav: ${speech.durationSeconds} s, '
'${speech.creditsConsumed} credits (request ${speech.requestId})');
// Transcribe it back, as an async job.
final job = await client.speechToText.jobs.create(
AudioFile.fromBytes(audio, 'speech.wav'),
language: 'ar',
);
final done = await client.speechToText.jobs.wait(job.jobId);
stdout.writeln('transcript: ${done.transcription}');
stdout.writeln(await client.speechToText.jobs
.export(done.id, format: ExportFormat.srt));
// Clean up.
await client.speechToText.jobs.delete(done.id);
await client.textToSpeech.jobs.delete(speech.jobId);
} on APIConnectionException catch (error) {
// The response was lost. When mayHaveExecuted is true, the request may
// have run (and been charged): look before sending it again.
stdout.writeln('no response: ${error.message}');
} on EchoNovaException catch (error) {
stdout.writeln('failed: $error');
exitCode = 1;
} finally {
client.close();
}
}
More examples #
Run them from the dart/ folder of the repository. Most need a key in ECHONOVA_API_KEY (proxy_mode.dart needs your proxy instead); each file starts with its exact command.
| Example | What it shows |
|---|---|
echonova_example.dart |
The example above: a voice, speech, a transcription job and an SRT export |
tts_generate.dart |
Text to Speech: pick a voice, synthesize, save the WAV |
tts_stream.dart |
Text to Speech over a stream: audio while it is generated, and speakStream for text that arrives in pieces |
stt_transcribe.dart |
Speech to Text of a file, with terminology and speaker labels |
stt_stream.dart |
Speech to Text over a stream: transcripts while the audio is still coming |
sound_cleaning.dart |
Sound Cleaning: clean a recording, save it, delete the job |
error_handling.dart |
Handling each kind of error, and checking for a job when a response was lost |
proxy_mode.dart |
EchoNova.proxy for Flutter apps: your backend holds the key |
agent_session.dart |
Starting, checking and ending an agent session from your server |