sherpa-onnx/dart-api-examples/README.md

Dart API Examples

This directory contains examples for the sherpa_onnx Dart API.

All examples can also be used in Flutter apps, even though they are pure Dart CLI programs here.

Start here: Read the version/ example first. It demonstrates all initialization patterns (sync, async, and isolate) that every other example depends on.

For Flutter users: Read hello_world for initialization, then vad-from-microphone or vad-from-file for how to copy model files from assets to a writable directory and build a complete app.

Initialization

All examples use the same simple initialization — no extra files needed:

import 'package:sherpa_onnx/sherpa_onnx.dart' as sherpa_onnx;

// Sync (Flutter and Dart CLI):
sherpa_onnx.initBindings();

// Or async (Flutter, Dart CLI, and web):
await sherpa_onnx.initBindingsAsync();

No path argument is required. The library auto-resolves the native library location for both Flutter apps and pure Dart CLI programs.

Isolates: If you use Dart isolates, you must call initBindings() or initBindingsAsync() in every isolate that uses sherpa-onnx. See version/ for examples.

Using in Flutter apps

These examples use model files from disk directly (e.g., ./model.onnx). In Flutter apps, model files must be bundled as assets and copied to a writable location before use, because Flutter apps run in a sandbox and cannot access arbitrary file paths.

Steps to use a model in Flutter:

1. Add the model file to your pubspec.yaml:

   flutter:
     assets:
       - assets/model.onnx
   

2. Copy the model from the asset bundle to a writable directory at runtime:

   import 'dart:io';
   import 'package:flutter/services.dart' show rootBundle;
   import 'package:path_provider/path_provider.dart';
   
   Future<String> copyAsset(String assetPath, String fileName) async {
     final dir = await getApplicationDocumentsDirectory();
     final file = File('${dir.path}/$fileName');
     if (!await file.exists()) {
       final data = await rootBundle.load(assetPath);
       await file.writeAsBytes(data.buffer.asUint8List());
     }
     return file.path;
   }
   

3. Pass the copied file path to the sherpa-onnx config:

   final modelPath = await copyAsset('assets/model.onnx', 'model.onnx');
   final config = sherpa_onnx.OfflineRecognizerConfig(
     model: sherpa_onnx.OfflineModelConfig(
       paraformer: sherpa_onnx.OfflineParaformerModelConfig(model: modelPath),
     ),
   );
   

For concrete Flutter examples, see flutter-examples/ in the repository.

Examples

Directory	Description
version	Version info — demonstrates sync, async, and isolate initialization
vad	Voice activity detection
vad-with-non-streaming-asr	VAD with non-streaming speech recognition (useful for subtitles)
non-streaming-asr	Non-streaming (offline) speech recognition
streaming-asr	Streaming (online) speech recognition
tts	Text to speech
speaker-diarization	Speaker diarization
speaker-identification	Speaker identification and verification
spoken-language-identification	Spoken language identification
audio-tagging	Audio tagging
keyword-spotter	Keyword spotting
add-punctuations	Adding punctuations to text
speech-enhancement-gtcrn	Speech enhancement/denoising with GTCRN
speech-enhancement-dpdfnet	Speech enhancement/denoising with DPDFNet (16 kHz family)
streaming-speech-enhancement-gtcrn	Streaming speech enhancement with GTCRN
streaming-speech-enhancement-dpdfnet	Streaming speech enhancement with DPDFNet

Running an example

cd vad
dart pub get
dart run ./bin/vad.dart --help

Creating a new example

dart create my-example
cd my-example

# Add sherpa_onnx to pubspec.yaml:
#   dependencies:
#     sherpa_onnx: ^1.13.5
#     path: ^1.9.0

dart pub get

In your main.dart:

import 'package:sherpa_onnx/sherpa_onnx.dart' as sherpa_onnx;

void main() async {
  await sherpa_onnx.initBindingsAsync();
  // Use sherpa-onnx APIs here...
}