Unofficial. This plugin is not affiliated with, endorsed by, sponsored by, or officially connected to Meta Platforms, Inc. "Meta", "Ray-Ban Meta", "Oakley Meta", and "Ray-Ban Display" are trademarks of their respective owners.

meta_wearables_dat_flutter

pub package License: MIT style: very good analysis Flutter

A Flutter plugin for Meta's Wearables Device Access Toolkit (DAT) on iOS and Android. It links Meta's official native SDKs and exposes one Dart API for Ray-Ban Meta, Oakley Meta, and Meta Ray-Ban Display glasses: registration, permissions, device state, camera streaming into a Flutter Texture, photo capture, on-glasses Display UI, the Mock Device Kit, and Meta's experimental modules.

It is a bridge, not a reimplementation. Meta's SDKs are closed-source binaries that this plugin depends on.

Status
Plugin version 1.0.0
Meta DAT version 1.0.0 (Meta's first stable, supported release)
Distribution Developer Mode and the invite-only Beta release channel in the Wearables Developer Center. Meta does not yet support publishing DAT apps to the App Store.
Developer Center Create a new app version for builds made with DAT 1.0.
Verification 1.0.0 is verified against Meta's Mock Device Kit on the iOS Simulator and Android emulator, plus unit tests on every layer. The real-glasses matrix in doc/release_checklist.md is still open; report hardware findings as issues.

Features

Feature iOS Android Notes
Registration (app-initiated) Yes Yes startRegistration(); the plugin hands the callback URL to the SDK
Registration started from Meta AI Yes Yes registrationRequestStream()
Permissions (camera, microphone) Yes Yes requestPermission(Permission.camera)
Device list and live state Yes Yes Battery, charging, wear, hinge, thermal, link, compatibility
Camera preview, raw codec, Flutter texture Yes Yes No frames over the method channel
Camera preview, hvc1 codec, Flutter texture Yes Yes VideoToolbox on iOS, MediaCodec on Android
Opt-in per-frame stream Yes Yes videoFramesStream(), gated on listeners
Photo capture (JPEG, HEIC) Yes Yes capturePhoto()
Background streaming Yes Yes Audio session on iOS, foreground service on Android
Display (Meta Ray-Ban Display) Yes Yes Declarative component DSL with callbacks
Diagnostics Yes Yes Info.plist / manifest findings, resource ledger
Mock Device Kit Yes Yes Simulated glasses, camera feed, permissions, display
Experimental modules Yes Yes Inputs, Motion, Speech, voice invocations, high-res photo, in-stream audio

Requirements

Minimum
Flutter 3.44.0 (Dart 3.8)
iOS 17.2, Swift Package Manager only (CocoaPods is not supported for this plugin)
Xcode 26.4 (Meta's binaries are built with Swift 6.3)
Android minSdk 31, MainActivity extends FlutterFragmentActivity
Android SDK source Maven Central (no GitHub token)
Meta AI app V290
Glasses firmware V128

Turn on Developer Mode in the Meta AI app: Settings > App Info, tap the version number five times.

Install

dependencies:
  meta_wearables_dat_flutter: ^1.0.0

flutter pub get

iOS setup

  1. Set the deployment target to 17.2 in Xcode (Runner target > General > Minimum Deployments).
  2. Swift Package Manager is Flutter's default since 3.44. If you turned it off, run flutter config --enable-swift-package-manager.
  3. Add the keys below to ios/Runner/Info.plist. Replace myapp with your own scheme (letters, digits, +, -, .; no underscores).
<key>MWDAT</key>
<dict>
  <!-- Must end with "://". Meta AI appends the callback query to it. -->
  <key>AppLinkURLScheme</key>
  <string>myapp://</string>
  <!-- "0" in Developer Mode; your app id from the Developer Center otherwise. -->
  <key>MetaAppID</key>
  <string>0</string>
  <key>ClientToken</key>
  <string>developer-mode-placeholder</string>
  <key>TeamID</key>
  <string>$(DEVELOPMENT_TEAM)</string>
</dict>

<key>CFBundleURLTypes</key>
<array>
  <dict>
    <key>CFBundleURLSchemes</key>
    <array>
      <string>myapp</string>
    </array>
  </dict>
</array>

<key>LSApplicationQueriesSchemes</key>
<array>
  <string>fb-viewapp</string>
</array>

<key>UISupportedExternalAccessoryProtocols</key>
<array>
  <string>com.meta.ar.wearable</string>
</array>

<key>UIBackgroundModes</key>
<array>
  <string>bluetooth-central</string>    <!-- required -->
  <string>external-accessory</string>   <!-- required -->
  <string>bluetooth-peripheral</string> <!-- recommended -->
  <string>processing</string>           <!-- recommended -->
  <string>audio</string>                <!-- for enableBackgroundStreaming() -->
</array>

<key>NSBluetoothAlwaysUsageDescription</key>
<string>Connects to your glasses.</string>
<key>NSLocalNetworkUsageDescription</key>
<string>Finds and connects to your glasses over Wi-Fi.</string>
<key>NSBonjourServices</key>
<array>
  <string>_bonjour._tcp</string>
</array>
<!-- Needed when you declare the audio background mode or use Speech. -->
<key>NSMicrophoneUsageDescription</key>
<string>Uses the glasses microphone.</string>

DAMEnabled is obsolete in DAT 1.0; remove it. Apps that use a scene-based lifecycle (SceneDelegate) must forward incoming URLs to the plugin; see doc/registration_flow.md. Run MetaWearablesDat.dumpDiagnostics() to check the result.

Android setup

  1. Make MainActivity extend FlutterFragmentActivity:

    import io.flutter.embedding.android.FlutterFragmentActivity

class MainActivity : FlutterFragmentActivity()

  2. Set minSdk = 31 in android/app/build.gradle.kts.

  3. Add to android/app/src/main/AndroidManifest.xml:

<uses-permission android:name="android.permission.BLUETOOTH" />
<uses-permission android:name="android.permission.BLUETOOTH_CONNECT" />
<uses-permission android:name="android.permission.INTERNET" />

<application ...>
  <!-- "0" in Developer Mode; Developer Center values otherwise. -->
  <meta-data
      android:name="com.meta.wearable.mwdat.APPLICATION_ID"
      android:value="0" />
  <meta-data
      android:name="com.meta.wearable.mwdat.CLIENT_TOKEN"
      android:value="0" />

  <activity android:name=".MainActivity" android:launchMode="singleTop" ...>
    <!-- Registration callback from the Meta AI app. -->
    <intent-filter>
      <action android:name="android.intent.action.VIEW" />
      <category android:name="android.intent.category.BROWSABLE" />
      <category android:name="android.intent.category.DEFAULT" />
      <data android:scheme="myapp" />
    </intent-filter>
  </activity>
</application>

Meta's Android SDK now resolves from Maven Central. You do not need a GitHub token, a GitHub Packages repository, or DAM_ENABLED. The plugin merges the foreground-service permissions and the background-streaming service into your manifest.

Quick start

import 'package:flutter/widgets.dart';
import 'package:meta_wearables_dat_flutter/meta_wearables_dat_flutter.dart';

Future<int?> startCamera() async {
  // Android: BLUETOOTH_CONNECT and SDK initialisation. iOS: returns true.
  await MetaWearablesDat.requestAndroidPermissions();

  // Register once. The Meta AI app opens and calls back into your app.
  if (await MetaWearablesDat.getRegistrationState() !=
      RegistrationState.registered) {
    await MetaWearablesDat.startRegistration();
    await MetaWearablesDat.registrationStateStream()
        .firstWhere((s) => s == RegistrationState.registered);
  }

  // Camera permission is granted in the Meta AI app.
  final status = await MetaWearablesDat.requestPermission(Permission.camera);
  if (!status.isGranted) return null;

  // Returns a texture id for Texture(textureId: ...).
  return MetaWearablesDat.startStreamSession(
    config: const StreamSessionConfig(
      quality: StreamQuality.high, // 720 x 1280
      frameRate: StreamFrameRate.fps24,
      videoCodec: VideoCodec.raw,
    ),
  );
}

Widget preview(int textureId) => AspectRatio(
  aspectRatio: 9 / 16,
  child: Texture(textureId: textureId),
);

Watch devices and their state:

MetaWearablesDat.devicesStream().listen((devices) {
  for (final d in devices) {
    debugPrint('${d.name}: ${d.linkState.name}, battery ${d.batteryLevel}, '
        '${d.donState.name}, thermal ${d.thermalLevel.name}');
  }
});

// One device; emits the current snapshot first, then changes.
MetaWearablesDat.deviceStateStream(uuid).listen((device) { /* ... */ });

Accept registrations started from the Meta AI app:

MetaWearablesDat.registrationRequestStream().listen((request) {
  request.continueRegistration(); // or request.cancel(); expires after 5 min
});

Capture a photo and stop:

final photo = await MetaWearablesDat.capturePhoto(format: PhotoFormat.jpeg);
// photo.bytes, photo.format (the actual encoding)

await MetaWearablesDat.stopStreamSession(); // also releases the texture

More: doc/streaming.md, doc/frame_processing.md, doc/device_state.md.

Display (Meta Ray-Ban Display)

await MetaWearablesDat.startDisplaySession();

final warnings = await MetaWearablesDat.sendDisplayView(
  FlexBox(
    spacing: 12,
    padding: 24,
    children: [
      const DisplayText('Hello from Flutter', style: DisplayTextStyle.heading),
      const DisplayIcon(DisplayIconName.checkmark),
      DisplayButtonGroup(
        buttons: [
          DisplayButton(
            label: 'Done',
            onClick: MetaWearablesDat.stopDisplaySession,
          ),
        ],
      ),
    ],
  ),
);
// warnings lists values the SDK substituted; they are not fatal.

MetaWearablesDat.displayStateStream().listen((state) => debugPrint(state.name));

The canvas is 600 x 600. Every sendDisplayView replaces the whole view. The Back gesture on the glasses ends the display session. VideoPlayer plays HTTPS MP4 and must be the root view. Use DisplayNode.validate() to check a tree before sending it. Camera, display, and experimental capabilities share one device session. See doc/display_access.md.

Error handling

Every failure is a subclass of the sealed DatError. Each error has a category (for example STREAM_ERROR), a code (the case name, for example hingesClosed), a typed reason on the concrete subclasses, a message, and a recoveryAction.

try {
  await MetaWearablesDat.startStreamSession();
} on DatError catch (error) {
  final text = switch (error) {
    DeviceSessionError(reason: DeviceSessionErrorCase.noEligibleDevice) =>
      'Connect and wear your glasses.',
    StreamError(reason: StreamErrorCase.hingesClosed) => 'Open the glasses.',
    StreamError(reason: StreamErrorCase.thermalHot) => 'The glasses are hot.',
    PermissionError() => 'Allow camera access in the Meta AI app.',
    _ => '${error.category}/${error.code}: ${error.message}',
  };
  debugPrint(text);

  switch (error.recoveryAction) {
    case DatRecoveryAction.openFirmwareUpdate:
      await MetaWearablesDat.openFirmwareUpdate();
    case DatRecoveryAction.openDatGlassesAppUpdate:
      await MetaWearablesDat.openDatGlassesAppUpdate();
    default:
      break;
  }
}

Asynchronous errors arrive on registrationErrorStream(), deviceSessionErrorStream(), streamErrorStream(), and displayErrorStream().

Diagnostics

final diagnostics = await MetaWearablesDat.dumpDiagnostics();
debugPrint('${diagnostics.platform} plugin ${diagnostics.pluginVersion}, '
    'DAT ${diagnostics.sdkVersion}');
for (final finding in diagnostics.findings) {
  debugPrint('$finding\n  fix: ${finding.fix}');
}

findings lists configuration problems with a stable id, a severity, and a fix. Examples: appLinkUrlSchemeSuffix (scheme does not end with ://), appLinkUrlSchemeNotRegistered, externalAccessoryProtocol, backgroundMode.external-accessory, damEnabledIgnored, and developerModeCredentials (informational). diagnostics.errors returns only the blocking ones. resources counts native textures, listeners, and sessions; diagnostics.isIdle is true when everything is released.

Experimental APIs

Meta marks some DAT 1.0 modules as experimental. The plugin exposes them with @experimental and excludes them from semantic versioning.

Warning: apps that use experimental APIs can be built and tested in Developer Mode and Beta release channels, but cannot ship to production release channels.

Capability API
Inputs (touchpad, buttons, Meta Neural Band) startInputs, inputEventsStream, stopInputs
Motion (head motion samples) startMotion(samplingRate:), motionSamplesStream, stopMotion
Speech (on-device transcription) startSpeech, transcriptionStream, stopSpeech
Voice invocations ("Hey Meta") startVoiceInvocations, voiceInvocationsStream, stopVoiceInvocations
High-resolution photo captureHighResPhoto(resolution:, quality:), photoTransferProgressStream
In-stream audio StreamSessionConfig(audio: AudioStreamConfig()), audioFramesStream 
await MetaWearablesDat.startInputs();
MetaWearablesDat.inputEventsStream().listen((event) {
  if (event is NavInputEvent) debugPrint('nav ${event.direction.name}');
});

The main library exports the experimental types too. package:meta_wearables_dat_flutter/experimental.dart exports only the experimental types, so you can use it as a marker for experimental usage (the analyzer reports it as unnecessary next to the main import).

On Android you can drop the experimental AARs with this line in your app's android/gradle.properties:

mwdat.experimental=false

Experimental calls then throw a DatPluginError with category EXPERIMENTAL_NOT_LINKED (isExperimentalNotLinked is true). iOS always links the experimental modules. See doc/experimental.md.

Testing with the Mock Device Kit

await MetaWearablesDat.enableMockDevice();
final glasses =
    await MetaWearablesDat.pairMockGlasses(MockGlassesModel.rayBanMeta);

// A mock device appears in devicesStream() after power on and unfold.
await MetaWearablesDat.mockPowerOn(glasses.uuid);
await MetaWearablesDat.mockUnfold(glasses.uuid);
await MetaWearablesDat.mockDon(glasses.uuid);

await MetaWearablesDat.setMockCameraFeed(glasses.uuid, h265VideoPath);
await MetaWearablesDat.setMockPermission(
  Permission.camera,
  PermissionStatus.granted,
);

final textureId =
    await MetaWearablesDat.startStreamSession(deviceUUID: glasses.uuid);
// ...
await MetaWearablesDat.stopStreamSession();
await MetaWearablesDat.disableMockDevice();

You can also simulate battery, charging, thermal level, touchpad taps, and capture failures, and preview the display with startMockTestServer(). Up to three mock devices can be paired. Disable the Mock Device Kit before you ship. See doc/mock_device.md.

Known issues

  • Enumerating all Objective-C classes crashes on iOS 18 (Meta SDK issue). MWDATCore 1.0.0 weakly links types that exist only on iOS 26 (from the Network and WiFiAware frameworks). Any call to objc_copyClassList on iOS 17.2 to 18.x forces those classes to initialize and aborts the app with "Failed to look up symbolic reference". Normal plugin use is unaffected; the integration tests pass on iOS 18. Confirmed on iOS 18.0 and 18.5 simulators, not on iOS 26.5 or 27. Before shipping to iOS 18 users, check that no SDK in your app enumerates every class (some analytics, crash-reporting and dependency-injection libraries do), and run XCTest bundles on an iOS 26+ simulator.
  • Raw frames pause in the iOS background. Use VideoCodec.hvc1 and enableBackgroundStreaming() to keep streaming.
  • sendMockDisplayClick returns true, but Meta has not documented the identifier format, so it may not fire onClick.
  • Xcode build after flutter pub get can fail with "requires minimum platform version 17.2 ... but this target supports 15.0". Set the app's iOS deployment target to 17.2 and run flutter build ios --config-only once.

More in doc/troubleshooting.md.

Samples

Sample Shows
samples/camera_access Flutter port of Meta's Camera Access sample: registration, streaming, photo and frame capture, Mock Device Kit menu.
samples/display_access Flutter port of Meta's Display Access sample: tutorial screens on Ray-Ban Display with button groups and video.
samples/glasses_companion DAT 1.0 features in one app: live device state, update deep links, diagnostics findings, hvc1 preview, high-res photo, a display card, mock controls and the experimental modules.

Documentation

Guide Topic
getting_started.md Install, Info.plist, manifest, Developer Mode
registration_flow.md Registration, callback URLs, Meta-AI-initiated requests
streaming.md Texture preview, codecs, photos, background streaming
frame_processing.md videoFramesStream, pixel formats, costs
display_access.md Display DSL and callbacks
device_state.md Battery, wear, hinge, thermal, compatibility
mock_device.md Mock Device Kit
experimental.md Inputs, Motion, Speech, voice invocations
troubleshooting.md Common problems
production_checklist.md Before you distribute
release_checklist.md Releasing this plugin
migration_0.7_to_1.0.md Upgrading from 0.7.x

Meta's documentation: wearables.developer.meta.com/docs/develop/. Changes: CHANGELOG.md.

Compatibility

The plugin's major.minor version tracks Meta's DAT version. The patch number is the plugin's own.

Plugin Meta DAT (iOS and Android)
1.0.x 1.0.0
0.7.x 0.7.0

Upgrading from 0.7.x: doc/migration_0.7_to_1.0.md.

Developer terms and data collection

By using the Wearables Device Access Toolkit you agree to the Meta Wearables Developer Terms and the Acceptable Use Policy. Meta may collect information about how users' devices communicate with your app. To opt out of analytics, add Analytics > OptOut = true inside the MWDAT dict on iOS, and <meta-data android:name="com.meta.wearable.mwdat.ANALYTICS_OPT_OUT" android:value="true" /> on Android.

License and maintainer

MIT. Copyright 2026 iSee Labs. Maintained by Talha Ordukaya. Issues: github.com/iSee-Labs/meta-wearables-dat-flutter/issues.

Meta's SDKs (iOS, Android) are linked as dependencies, not redistributed, and remain under Meta's license terms. See NOTICE.

"Meta", "Ray-Ban Meta", "Oakley Meta", and "Ray-Ban Display" are trademarks of Meta Platforms, Inc. and/or its affiliates. This project is not affiliated with Meta.

Libraries

experimental
The experimental DAT 1.0 capabilities of meta_wearables_dat_flutter.
meta_wearables_dat_flutter
Unofficial Flutter plugin for Meta's Wearables Device Access Toolkit (DAT) 1.0 on iOS and Android. Not affiliated with Meta Platforms, Inc.