tflite_flutter_custom

A drop-in alternative to tflite_flutter, made easier. tflite_flutter_custom automatically includes all required native TensorFlow Lite libraries for Android, iOS, macOS, Windows, and Linux.

No manual setup or handling of native libraries required.

Why this fork?

tflite_flutter is an excellent library, but it requires developers to manually include platform-specific native binaries.
This fork removes that friction by bundling everything inside the package.

• No more copying .so, .dll, or .dylib files
• Works out of the box on all supported platforms
• Same API as tflite_flutter

Installation

dependencies:
  tflite_flutter_custom: ^1.0.2

Usage

The API is identical to tflite_flutter, so you can follow their documentation directly. For example:

import 'package:tflite_flutter_custom/tflite_flutter_custom.dart';

void main() async {
  final interpreter = await Interpreter.fromAsset('model.tflite');
  print('Model loaded successfully!');
}

Platform Support

• ✅ Android
• ✅ iOS
• ✅ macOS
• ✅ Windows
• ✅ Linux

All required native binaries are automatically included in the build.

Credits

This project is based on tflite_flutter by the TensorFlow team and contributors. All credit for the original bindings goes to them.