Skip to content

Conversation

@DeepBhupatkar
Copy link

This PR introduces the VideoSDK.live Audio Transport into the pipecat-ai/pipecat framework, allowing Pipecat-ai bots to join VideoSDK.live meetings, exchange realtime audio with participants, and respond to participant chat messages via VideoSDK PubSub.

1.Added Dependency :

  • Add the VideoSDK.live transport implementation dependency
  • Updated pyproject.toml to include the VideoSDK Python SDK:
    videosdk = ["videosdk>=0.2.7"]

2.Environment Variables :

  • Added new VideoSDK.live environment variables to env.example:
# VideoSDK.live
VIDEOSDK_AUTH_TOKEN=...
VIDEOSDK_MEETING_ID=...

3.VideoSDK's Transport Implementation :

  • Added a new videosdk folder inside the transports directory to organize all VideoSDK.live transport components.
src/pipecat/transports/videosdk/
├── __init__.py        # placeholder
└── transport.py       # VideoSDKTransport implementation

Features include:

  • Realtime bidirectional audio (incoming & outgoing)
  • Meeting join and authentication logic
  • Audio pipeline integration with Pipecat
  • Support for VideoSDK.live meeting and functinoality like chat.

4.New Example Added :

  • Added a full usage example demonstrating VideoSDK.live audio transport:
    examples/foundational/04c-transports-videosdk.py
  • This example shows how to initialize the transport and run a complete STT → LLM → TTS audio pipeline.

DeepBhupatkar and others added 5 commits November 26, 2025 18:11
Introduces full support for VideoSDK.live meetings, allowing bots to send/receive audio and process chat messages from participants.

- Implement `VideoSDKTransport` for bidirectional audio streaming.
- Add `VideoSDKParams` for configuration of meeting credentials and audio settings.
- Add `PipecatAudioTrack` to handle custom audio output and synchronization.
- Integrate VideoSDK PubSub to enable text-to-voice interactions via `on_pubsub_message_received`.
- Adds `VIDEOSDK_AUTH_TOKEN` and `VIDEOSDK_MEETING_ID` to support the new VideoSDK transport integration.
- Initializing the transport with environment variables.
- Setting up a full audio pipeline (STT, LLM, TTS).
- Handling bidirectional audio and PubSub chat messages.
@DeepBhupatkar
Copy link
Author

DeepBhupatkar commented Nov 26, 2025

Hi @markbackman,@filipi87 & @aconchillo - This PR is ready for review.Please review when possible.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant