Godot-LiveKit

Godot-LiveKit is a GDExtension for Godot 4.5 that integrates the LiveKit C++ SDK, allowing you to build real-time voice, video, and data applications directly within the Godot Engine using GDScript or C#.

Features

• Real-Time Communication: Connect to LiveKit servers for audio, video, and data streaming.
• Full Track Support: Publish and subscribe to audio/video tracks, with local sources for capturing from Godot.
• Screen Capture: Capture monitors or individual windows natively using the built-in LiveKitScreenCapture class (macOS, Windows, Linux).
• Data Channels: Send and receive arbitrary data messages with reliable or unreliable delivery.
• RPC Support: Perform remote procedure calls between participants.
• End-to-End Encryption (E2EE): Secure your media streams with configurable encryption, key management, and per-participant frame cryptors.
• Connection Statistics: Access detailed WebRTC statistics including inbound/outbound RTP, codecs, transport, and candidate pair metrics.
• Fast Build Times: Uses prebuilt binaries for both godot-cpp and the LiveKit C++ SDK, reducing compilation time to seconds instead of hours.
• Cross-Platform: Supports Linux, macOS (Universal), and Windows.
• Native GDExtension: Works out-of-the-box with Godot 4.5 without requiring custom engine builds.

Prerequisites

To build the extension from source, you will need:

• Python 3.x
• SCons (pip install scons)
• CMake (Optional, for advanced builds)

Platform-Specific Requirements:

• Linux: g++, curl, tar, unzip
• macOS: Xcode Command Line Tools, curl, tar, unzip
• Windows: MSVC (Visual Studio Build Tools) or MinGW-w64 (if cross-compiling from Linux/macOS)

Building from Source

This repository includes a custom build.sh script that automatically fetches the required LiveKit C++ SDK and godot-cpp prebuilt binaries, and compiles the extension.

1. Clone the repository:

   git clone https://github.com/NodotProject/godot-livekit.git
   cd godot-livekit

2. Run the build script for your platform:

   • Linux:

     ./build.sh linux

   • macOS:

     ./build.sh macos

   • Windows:

     ./build.sh windows

Once the build is complete, the compiled dynamic libraries (.so, .dylib, or .dll) and the LiveKit shared libraries will be placed in the addons/godot-livekit/bin/ directory.

Usage in Godot

1. Copy the addons/godot-livekit folder into your Godot project's addons/ directory.
2. Enable the plugin from the Godot Editor: Project -> Project Settings -> Plugins.
3. You can now access LiveKit classes directly from GDScript:

extends Node

var room: LiveKitRoom

func _ready():
    room = LiveKitRoom.new()
    room.connected.connect(_on_connected)
    room.participant_connected.connect(_on_participant_connected)
    room.track_subscribed.connect(_on_track_subscribed)
    room.data_received.connect(_on_data_received)
    room.connect_to_room("wss://your-server.url", "your-token", {})

# No _process() needed — rooms, video streams, and screen captures are
# auto-polled every frame by default. Set auto_poll = false if you prefer
# to call poll_events() / poll() manually.

func _on_connected():
    print("Connected as: ", room.get_local_participant().get_identity())

func _on_participant_connected(participant):
    print("Joined: ", participant.get_identity())

func _on_track_subscribed(track, publication, participant):
    print("Track subscribed: ", track.get_name())

func _on_data_received(data, participant, kind, topic):
    print("Data from ", participant.get_identity(), ": ", data.get_string_from_utf8())

Screen Capture

You can capture screens and windows natively:

var capture = LiveKitScreenCapture.create()
capture.frame_received.connect(_on_frame)
capture.start()

func _on_frame():
    var image = capture.get_image()
    # Feed into a LiveKitVideoSource for screen sharing

Running Tests

If you have GUT (Godot Unit Test) installed in addons/gut, you can run the test suite using the provided test.sh script:

./test.sh

Continuous Integration

The project is configured with GitHub Actions to automatically build releases for Linux, Windows, and macOS whenever a new tag (e.g., v1.0.0) is pushed. Check the .github/workflows directory for details.

License

This project is licensed under the MIT License. See the LICENSE file for more information. Note that the LiveKit C++ SDK is licensed under the Apache License 2.0.