MLKIT Image Labeler

Plugin which provides native ML Kit ImageLabeler API



  • Set minSdkVersion 21 in android/app/build.gradle

  • Add <meta-data android:name="" android:value="ica" /> in android/src/main/AndroidManifest.xml

  • Note: In case you are using multiple models separate them with commas android:value="ica,ocr"

  • Set ext.kotlin_version = '1.6.10' in android/build.gradle

  • App size impact: 700KB(Using Unbundled Model, downloads model using Google play services for the first time when app launches), refer here


  • Minimum iOS Deployment Target: 10.0
  • Xcode 12.5.1 or greater.
  • ML Kit only supports 64-bit architectures (x86_64 and arm64). Check this list to see if your device has the required device capabilities.
  • Since ML Kit does not support 32-bit architectures (i386 and armv7) Read more, you need to exclude amrv7 architectures in Xcode in order to build iOS, refer here


// Create an Instance of [MlKitImageLabeler]
final imageLabeler = MlKitImageLabeler();

// Pick Image using image picker

// Call `processImage()` and pass params as `InputImage` [check example for more info]
final labels = await imageLabeler.processImage(

String predictedLabels = '';

// Iterate over Labels 
for (var label in labels) {
    predictedLabels +=
        '\n${label.label} ${(label.confidence * 100).toStringAsPrecision(2)}%';

This plugin is basically a trimmed down version of google_ml_kit. As google_ml_kit contains all the NLP and Vison APIs, the App size increases drastically. So, I created this plugin and now the example app's fat apk is of 17MB and splitted apks are 6MB.