Skip to main content

Voice mode

The Langflow Playground supports voice mode for interacting with your applications through a microphone.

An OpenAI API key is required to use voice mode. An ElevenLabs API key enables more voices in the chat, but is optional.

Your flow must have a Chat input component to interact with the Playground.

Prerequisite​

Use voice mode in the Langflow Playground​

Chat with an agent in the Playground, and get more recent results by asking the agent to use tools.

  1. Create a Simple agent starter project.
  2. Add your OpenAI API key credentials to the Agent component.
  3. To start a chat session, click Playground.
  4. To enable voice mode, click the icon. The Voice mode pane opens.
  5. In the OpenAI API Key field, add your OpenAI API key credentials. This key is saved as a global variable in Langflow and is accessible from any component or flow.
  6. Your browser may prompt you for microphone access. Browser access is required to use voice mode. To continue, allow microphone access in your browser.
  7. In the Audio Input menu, select the input device to use with voice mode.
tip

A higher quality microphone improves OpenAI's voice chat comprehension.

  1. Optionally, add your ElevenLabs API key in the ElevenLabs API Key field. This makes more voices available for your AI responses. This key is saved as a global variable in Langflow and is accessible from any component or flow.
  2. In the Preferred Language menu, select your language for conversing with Langflow. This option changes both the spoken conversation and the chat responses in the Playground.
  3. Talk into your microphone. The waveform in the voice mode pane should register your input, and the agent should respond in voice and in the Playground.
  4. Ask the agent to use the tools available to find recent news about a subject.

The agent describes its search process, including accessing the URL tool to fetch recent news. The agent summarizes the recent news in speech and in the Playground.

Be aware of the following considerations when using voice mode:

  • Name and describe your tools accurately, so the Agent chooses tools correctly.
  • Voice mode does not use the instructions in the Agent component's Agent Instructions field, because your spoken instructions override this value.
  • Voice mode only maintains context within the conversation session you are currently in. If you exit a conversation and close the Playground, your conversational context is not available in the next chat session.

Langflow voice mode endpoints​

Langflow exposes OpenAI Realtime API-compatible websocket endpoints for your flows. You can build voice applications against these endpoints the same way you would build against OpenAI Realtime API websockets.

The WebSockets endpoints require an OpenAI API key for authentication, and they support an optional ElevenLabs integration.

Langflow exposes two WebSockets endpoints:

  • /ws/flow_as_tool/{flow_id} or /ws/flow_as_tool/{flow_id}/{session_id}: Establishes a connection to OpenAI Realtime voice, and then invokes flows as tools by the OpenAI Realtime model. This approach is ideal for low latency applications, but it is less deterministic since the OpenAI voice-to-voice model determines when to call your flow.

  • /ws/flow_tts/{flow_id} or /ws/flow_tts/{flow_id}/{session_id}: Converts audio to text using OpenAI Realtime voice transcription, and then each flow is invoked directly for each transcript. This approach is more deterministic but has higher latency. This is the mode used in the Langflow playground.

Path parameters:

  • flow_id: Required path parameter. The ID of the flow to be used as a tool.
  • session_id: Optional path parameter. A unique identifier for the conversation session. If not provided, one is automatically generated.
Search