Flutter Anam SDK

A Flutter package for integrating Anam AI's real-time avatar system into your Flutter applications. Built on WebRTC, this is a Flutter port of the Anam JavaScript SDK.

Features

Real-time avatar video/audio streaming via WebRTC
Two-way audio communication with microphone controls
Text messaging with full message history
Event-driven architecture for reactive UIs
Custom persona configuration (avatar, voice, system prompt, language)
Server-side session management with WebSocket proxy support
Custom audio mode for bring-your-own-transcription pipelines
Built-in AnamAvatarView widget
Cross-platform: iOS, Android, macOS, Web

Installation

Add to your pubspec.yaml:

dependencies:
  flutter_anam_sdk: ^0.0.1

Then run:

flutter pub get

Platform Setup

iOS

Add to ios/Runner/Info.plist:

<key>NSMicrophoneUsageDescription</key>
<string>Required for voice communication with the avatar</string>
<key>NSCameraUsageDescription</key>
<string>Required for video communication with the avatar</string>

macOS

Add to macos/Runner/Info.plist:

<key>NSMicrophoneUsageDescription</key>
<string>Required for voice communication with the avatar</string>
<key>NSCameraUsageDescription</key>
<string>Required for video communication with the avatar</string>

Add to both macos/Runner/DebugProfile.entitlements and macos/Runner/Release.entitlements:

<key>com.apple.security.network.client</key>
<true/>
<key>com.apple.security.device.audio-input</key>
<true/>
<key>com.apple.security.device.camera</key>
<true/>

Android

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

<uses-permission android:name="android.permission.INTERNET" />
<uses-permission android:name="android.permission.RECORD_AUDIO" />
<uses-permission android:name="android.permission.MODIFY_AUDIO_SETTINGS" />
<uses-permission android:name="android.permission.ACCESS_NETWORK_STATE" />
<uses-permission android:name="android.permission.CAMERA" />

Quick Start

1. Create a client

Production (session token from your server):

import 'package:flutter_anam_sdk/flutter_anam_sdk.dart';

final client = AnamClientFactory.createClient(
  sessionToken: 'your-session-token',
);

Development only (direct API key — never ship to production):

final client = AnamClientFactory.unsafeCreateClientWithApiKey(
  apiKey: 'your-api-key',
  enableLogging: true,
);

2. Set up a video renderer

import 'package:flutter_webrtc/flutter_webrtc.dart';

final renderer = RTCVideoRenderer();
await renderer.initialize();

3. Start a session

final personaConfig = PersonaConfig(
  personaId: 'your-persona-id',
  name: 'AI Assistant',
  avatarId: 'default_avatar',
  voiceId: 'default_voice',
  systemPrompt: 'You are a helpful assistant.',
);

await client.talk(
  personaConfig: personaConfig,
  onStreamReady: (stream) {
    if (stream != null) {
      renderer.srcObject = stream;
    }
  },
);

4. Display the avatar

AnamAvatarView(
  renderer: renderer,
  onMicToggle: () => client.setInputAudioEnabled(!client.inputAudioEnabled),
  isMicEnabled: client.inputAudioEnabled,
)

5. Send messages and control the session

// Send a text message
client.sendUserMessage('Hello!');

// Interrupt the avatar while it speaks
client.interruptPersona();

// Mute / unmute microphone
client.muteInputAudio();
client.unmuteInputAudio();

// End the session
await client.stopStreaming();

// Clean up when done
client.dispose();

Event Handling

Subscribe to events with client.on<T>(AnamEvent.xxx) which returns a Stream<T>:

// Message history updates
client.on<List<Message>>(AnamEvent.messageHistoryUpdated).listen((messages) {
  setState(() => _messages = messages);
});

// Connection lifecycle
client.on(AnamEvent.connectionEstablished).listen((_) {
  print('WebRTC connected');
});

client.on(AnamEvent.connectionClosed).listen((_) {
  print('Session ended');
});

// Stream events
client.on(AnamEvent.videoStreamStarted).listen((_) {
  print('Video stream is live');
});

client.on(AnamEvent.sessionReady).listen((_) {
  print('Avatar session is ready');
});

// Avatar state
client.on(AnamEvent.personaTalking).listen((_) {
  print('Avatar is speaking');
});

client.on(AnamEvent.personaListening).listen((_) {
  print('Avatar is listening');
});

// Errors and warnings
client.on(AnamEvent.error).listen((error) {
  print('Error: $error');
});

client.on(AnamEvent.warning).listen((warning) {
  print('Warning: $warning');
});

Available Events

Event	Payload	Description
`messageHistoryUpdated`	`List<Message>`	Message history changed
`connectionEstablished`	—	WebRTC connected
`connectionClosed`	—	Session ended
`videoStreamStarted`	`MediaStream`	Video stream is active
`audioStreamStarted`	—	Audio stream is active
`sessionReady`	`Map`	Avatar session is ready
`personaTalking`	—	Avatar began speaking
`personaListening`	—	Avatar is now listening
`inputAudioEnabled`	—	Microphone enabled
`inputAudioDisabled`	—	Microphone disabled
`error`	`dynamic`	An error occurred
`warning`	`dynamic`	A non-fatal warning

Custom Audio (Bring Your Own Transcription)

Disable the client microphone to implement your own speech recognition, VAD, or custom turn-taking logic. The avatar's audio is still received and played; only outbound mic capture is disabled.

final client = AnamClientFactory.createClient(
  sessionToken: sessionToken,
  disableClientAudio: true,
);

await client.talk(
  personaConfig: personaConfig,
  onStreamReady: (stream) { /* ... */ },
);

With disableClientAudio: true:

No microphone permission is requested
Audio transceiver is set to recvonly
Data channel remains available for sending messages

Send audio chunks directly to the avatar's lip-sync engine (Brain-less mode):

// Send PCM audio chunk (pcm_s16le, 24 kHz, mono)
client.appendInputAudio(Uint8List audioBytes);

// Signal end of a speech turn
client.endAgentAudioSequence();

API Reference

`AnamClientFactory`

Method	Description
`createClient({sessionToken, enableLogging, disableClientAudio})`	Create a client with a session token (recommended)
`createClientWithOptions({sessionToken, personaId, ...})`	Create a client with a default persona baked in
`unsafeCreateClientWithApiKey({apiKey, enableLogging})`	Development only — exposes your API key
`createWebRTCOffer({iceServers, disableClientAudio})`	Generate a WebRTC offer without a full client (advanced)

`AnamClient`

Methods

Method	Description
`talk({personaConfig, onStreamReady, preNegotiatedSession, proxyUrl})`	Start an avatar session
`stopStreaming([reason])`	End the current session
`sendUserMessage(content)`	Send a text message to the avatar
`interruptPersona()`	Interrupt the avatar while it is speaking
`setInputAudioEnabled(bool)`	Enable or disable the microphone
`muteInputAudio()`	Mute the microphone
`unmuteInputAudio()`	Unmute the microphone
`appendInputAudio(Uint8List)`	Stream raw PCM audio to the avatar (Brain-less mode)
`endAgentAudioSequence()`	Signal end of a speech turn (Brain-less mode)
`streamToVideoElement(renderer)`	Attach the remote stream to a renderer
`on<T>(AnamEvent)`	Subscribe to an event, returns `Stream<T>`
`dispose()`	Stop the session and release all resources

Properties

Property	Type	Description
`isSessionActive`	`bool`	Whether a session is currently running
`isConnected`	`bool`	Whether WebRTC is connected
`isDataChannelOpen`	`bool`	Whether the data channel is open
`inputAudioEnabled`	`bool`	Whether the microphone is active
`messageHistory`	`List<Message>`	Read-only message history
`currentPersona`	`PersonaConfig?`	Active persona configuration
`currentSessionId`	`String?`	Active session ID

`PersonaConfig`

PersonaConfig({
  required String personaId,         // Anam persona ID
  required String name,              // Display name
  required String avatarId,          // Avatar asset ID
  required String voiceId,           // Voice asset ID
  String? llmId,                     // Override the default LLM
  String? systemPrompt,              // System prompt for the avatar
  int? maxSessionLengthSeconds,      // Session timeout
  String? languageCode,              // e.g. 'en-US'
})

`AnamAvatarView`

AnamAvatarView({
  required RTCVideoRenderer? renderer,  // Video renderer (null shows placeholder)
  bool showControls = true,             // Show the mic toggle button
  VoidCallback? onMicToggle,            // Callback for the mic button
  bool isMicEnabled = true,             // Current mic state (controls button icon)
  double borderRadius = 12.0,           // Corner radius
  Color backgroundColor = Colors.black, // Background color
})

`Message`

class Message {
  final String id;
  final MessageRole role;    // MessageRole.user | MessageRole.assistant
  final String content;
  final DateTime timestamp;
}

Contributing

Contributions are welcome! Please open an issue or submit a pull request on GitHub.

License

This project is licensed under the MIT License — see the LICENSE file for details.

Built with ❤️ by DevCodeSpace

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
example		example
lib		lib
screenshots		screenshots
.flutter-plugins-dependencies		.flutter-plugins-dependencies
CHANGELOG.md		CHANGELOG.md
IMPLEMENTATION_SUMMARY.md		IMPLEMENTATION_SUMMARY.md
LICENSE		LICENSE
PUBLISHING.md		PUBLISHING.md
README.md		README.md
README_SERVER_NEGOTIATION.md		README_SERVER_NEGOTIATION.md
SERVER_SIDE_SESSION_MANAGEMENT.md		SERVER_SIDE_SESSION_MANAGEMENT.md
analysis_options.yaml		analysis_options.yaml
pubspec.lock		pubspec.lock
pubspec.yaml		pubspec.yaml

Navigation Menu

Repository files navigation

Flutter Anam SDK

Features

Installation

Platform Setup

iOS

macOS

Android

Quick Start

1. Create a client

2. Set up a video renderer

3. Start a session

4. Display the avatar

5. Send messages and control the session

Event Handling

Available Events

Custom Audio (Bring Your Own Transcription)

API Reference

`AnamClientFactory`

`AnamClient`

Methods

Properties

`PersonaConfig`

`AnamAvatarView`

`Message`

Contributing

License

About

Releases

Packages

Contributors

Languages

Folders and files

Latest commit

History

Repository files navigation

Flutter Anam SDK

Features

Installation

Platform Setup

iOS

macOS

Android

Quick Start

1. Create a client

2. Set up a video renderer

3. Start a session

4. Display the avatar

5. Send messages and control the session

Event Handling

Available Events

Custom Audio (Bring Your Own Transcription)

API Reference

AnamClientFactory

AnamClient

Methods

Properties

PersonaConfig

AnamAvatarView

Message

Contributing

License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

`AnamClientFactory`

`AnamClient`

`PersonaConfig`

`AnamAvatarView`

`Message`