executorch_dart library

ExecuTorch on-device ML inference for pure Dart.

Runs ExecuTorch models on Android, iOS, macOS, Linux, and Windows through dart:ffi and native assets. Works in any Dart program, including servers and command-line tools — no Flutter required.

import 'package:executorch_dart/executorch_dart.dart';

final model = await ExecuTorchModel.load('/path/to/model.pte');
final outputs = await model.forward([inputTensor]);
await model.dispose();

Flutter applications should depend on executorch_flutter instead, which adds asset-bundle loading and web support on top of this package.