Pipecat API Reference

https://reference-server.pipecat.ai/en/latest
Admin
This is the API reference for Pipecat, a platform for building real-time data pipelines. It provides...
Tokens:300,605
Snippets:2,654
Trust Score:7.5
Update:4 months ago
Show doc for...
Tokens:
Raw
### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.llm_service.html

Starts the LLM service.

```APIDOC
## start

### Description
Start the LLM service.

### Parameters
* **frame** (StartFrame) - The start frame.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.elevenlabs.stt.html

Starts the STT service and establishes a WebSocket connection.

```APIDOC
## start(_frame: StartFrame_)

### Description
Start the STT service and establish WebSocket connection.

### Parameters
* **frame** (StartFrame) - Frame indicating service should start.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.nvidia.stt.html

Starts the NVIDIA Nemotron Speech STT service and initializes streaming configuration.

```APIDOC
## start

### Description
Start the NVIDIA Nemotron Speech STT service and initialize streaming configuration.

### Parameters
* **frame** (StartFrame_) - StartFrame indicating pipeline start.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.minimax.tts.html

Starts the MiniMax TTS service, preparing it for operation with the given initialization parameters.

```APIDOC
## start

### Description
Start the MiniMax TTS service.

### Parameters
* **frame** (StartFrame_) - Required - The start frame containing initialization parameters.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.nvidia.tts.html

Starts the NVIDIA Nemotron Speech TTS service.

```APIDOC
## start(_frame: StartFrame)

### Description
Start the NVIDIA Nemotron Speech TTS service.

### Parameters
* **frame** (StartFrame) - The start frame for the service.
```

--------------------------------

### setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.asyncai.tts.html

Sets up the processor with required components, typically configuration objects.

```APIDOC
## setup

### Description
Set up the processor with required components.

### Parameters
- **setup** (FrameProcessorSetup) - Required - Configuration object containing setup parameters.
```

--------------------------------

### setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.inworld.tts.html

Sets up the TTS processor with the required components.

```APIDOC
## async def setup(_setup : FrameProcessorSetup_)

### Description
Set up the processor with required components.

### Parameters
* **setup** (FrameProcessorSetup_) - Configuration object containing setup parameters.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.aws.stt.html

Initializes the connection to AWS Transcribe when the service starts.

```APIDOC
## start

### Description
Initialize the connection when the service starts.

### Parameters
* **frame** (StartFrame_) - Start frame signaling service initialization.
```

--------------------------------

### setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.mistral.tts.html

Sets up the Mistral TTS processor with required components.

```APIDOC
## setup

### Description
Set up the processor with required components.

### Method
`async`

### Parameters
* **setup** (FrameProcessorSetup) - Configuration object containing setup parameters.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.azure.stt.html

Starts the speech recognition service with a given start frame.

```APIDOC
## async start(_frame: StartFrame_)

### Description
Start the speech recognition service.

### Parameters
* **frame** (StartFrame_) - Frame indicating the start of processing.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.audio.mixers.soundfile_mixer.html

Initializes the mixer and loads all sound files based on the provided sample rate.

```APIDOC
## start

### Description
Initialize the mixer and load all sound files.

### Parameters
* **sample_rate** (int) - The sample rate of the output transport in Hz.
```

--------------------------------

### setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.elevenlabs.tts.html

Sets up the processor with the required components for ElevenLabs TTS.

```APIDOC
## async setup(_setup : FrameProcessorSetup_) ##

### Description
Sets up the processor with the required components for ElevenLabs TTS.

### Parameters
* **setup** (FrameProcessorSetup) - Required - Configuration object containing setup parameters.
```

--------------------------------

### Serializer Setup Method

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.serializers.base_serializer.html

Initializes the serializer with startup configuration provided in a StartFrame.

```python
async setup(_frame : StartFrame_)[source]
    
Initialize the serializer with startup configuration. 

Parameters:
    
**frame** – StartFrame containing initialization parameters. 
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.inworld.tts.html

Start the Inworld WebSocket TTS service.

```APIDOC
## start

### Description
Start the Inworld WebSocket TTS service.

### Parameters
* **_frame** (StartFrame_) - The start frame.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.llm_service.html

Starts the LLM service and establishes the WebSocket connection. This method is triggered by a StartFrame.

```APIDOC
## start

### Description
Starts the LLM service and establishes the WebSocket connection. This method is triggered by a StartFrame.

### Parameters
* **frame** (StartFrame) - The start frame triggering service initialization.
```

--------------------------------

### Setup Serializer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.serializers.genesys.html

Sets up the serializer with the provided pipeline configuration.

```APIDOC
## Method setup

### Description
Sets up the serializer with pipeline configuration.

### Parameters
- **frame** (StartFrame_) - The StartFrame containing pipeline configuration.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.inworld.tts.html

Starts the Inworld TTS service.

```APIDOC
## async def start(_frame : StartFrame_)

### Description
Start the Inworld TTS service.

### Parameters
* **frame** (StartFrame_) - The start frame.
```

--------------------------------

### Setup ExotelFrameSerializer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.serializers.exotel.html

Sets up the serializer with pipeline configuration using a StartFrame.

```python
await serializer.setup(_frame = StartFrame())

```

--------------------------------

### start(_frame: StartFrame_)

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.gladia.stt.html

Initiates the Gladia STT websocket connection.

```APIDOC
## start(_frame: StartFrame_)

### Description
Start the Gladia STT websocket connection.

### Parameters
* **frame** (StartFrame_) - The start frame triggering service startup.
```

--------------------------------

### HeyGenClient.start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.html

Starts the HeyGen client, initiating the connection to the HeyGen service.

```APIDOC
## HeyGenClient.start

### Description
Starts the HeyGen client, initiating the connection to the HeyGen service.

### Method
`HeyGenClient.start()`

### Parameters
None

### Request Example
```python
client.start()
```

### Response
None. This method starts the client's operation.
```

--------------------------------

### HeyGenVideoService Start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.video.html

Starts the HeyGen video service and initializes the avatar session, creating tasks for audio/video processing.

```APIDOC
## start

### Description
Start the HeyGen video service and initialize the avatar session. Creates necessary tasks for audio/video processing and establishes the connection with the HeyGen service.

### Parameters
* **frame** (StartFrame_) - The start frame containing initialization parameters.
```

--------------------------------

### Start AIService

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.ai_service.html

Starts the AI service, initiating its processing lifecycle. Subclasses should override this to implement service-specific startup logic.

```python
async def start(_frame : StartFrame_):
    """Start the AI service."""
    pass
```

--------------------------------

### Multiple Transport Bot Example

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.run.html

An example demonstrating how to set up different transports based on the runner arguments type for supporting multiple protocols.

```python
async def bot(runner_args: RunnerArguments):
    # Type-safe transport detection
    if isinstance(runner_args, DailyRunnerArguments):
        transport = setup_daily_transport(runner_args)  # Your application code
    elif isinstance(runner_args, SmallWebRTCRunnerArguments):
        transport = setup_webrtc_transport(runner_args)  # Your application code
    elif isinstance(runner_args, WebSocketRunnerArguments):
        transport = setup_telephony_transport(runner_args)  # Your application code

    # Your bot implementation
    await run_pipeline(transport)


```

--------------------------------

### MistralSTTService.start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.mistral.stt.html

Starts the STT service and establishes a connection.

```APIDOC
## MistralSTTService.start

### Description
Starts the STT service and establishes the connection to the Mistral transcription service.

### Parameters
*   **frame** (StartFrame_) - Frame indicating the service should start.
```

--------------------------------

### setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.cartesia.tts.html

Sets up the Cartesia TTS processor with the necessary components and configuration.

```APIDOC
## async setup(_setup : FrameProcessorSetup_)

### Description
Set up the processor with required components.

### Parameters
#### Path Parameters
* **setup** (FrameProcessorSetup) - Required - Configuration object containing setup parameters.
```

--------------------------------

### AIService Start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.ai_service.html

Starts the AI service, initiating its processing lifecycle. Subclasses should override this to implement service-specific startup logic.

```APIDOC
## `AIService.start`

### Description
Starts the AI service. Called when the service should begin processing. Subclasses should override this method to perform service-specific initialization.

### Parameters
* **frame** (`StartFrame`) - The start frame containing initialization parameters.
```

--------------------------------

### Setup VonageFrameSerializer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.serializers.vonage.html

Sets up the serializer by processing the initial StartFrame containing pipeline configuration.

```python
async def setup(_frame : StartFrame_):
    
    Sets up the serializer with pipeline configuration. 

    Parameters:
        
    **frame** – The StartFrame containing pipeline configuration.
```

--------------------------------

### Start Recording Audio

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.processors.audio.audio_buffer_processor.html

Starts recording audio from both user and bot sources. This method initializes the recording state and resets any existing audio buffers.

```python
await processor.start_recording()
```

--------------------------------

### Setup Processor

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.neuphonic.tts.html

Sets up the TTS processor with the necessary components for operation.

```APIDOC
## setup

### Description
Set up the processor with required components.

### Parameters
* **setup** (FrameProcessorSetup_) - Required - Configuration object containing setup parameters.
```

--------------------------------

### Setup TTS Processor

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.deepgram.tts.html

Sets up the TTS processor with the required components and configuration.

```APIDOC
## Setup TTS Processor

### Description
Sets up the TTS processor with the required components and configuration.

### Parameters
* **setup** (FrameProcessorSetup) - Required - Configuration object containing setup parameters.
```

--------------------------------

### HeyGenClient.setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.html

Sets up the HeyGen client with initial configuration, including service type and callbacks.

```APIDOC
## HeyGenClient.setup

### Description
Sets up the HeyGen client with initial configuration, including service type and callbacks.

### Method
`HeyGenClient.setup(service_type: ServiceType, callbacks: HeyGenCallbacks)`

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **service_type** (ServiceType) - Required - The type of HeyGen service to use (e.g., `INTERACTIVE_AVATAR`, `LIVE_AVATAR`).
- **callbacks** (HeyGenCallbacks) - Required - An object containing callback functions for different events.
  - **on_connected** (callable) - Optional - Callback when the client connects.
  - **on_participant_connected** (callable) - Optional - Callback when a participant connects.
  - **on_participant_disconnected** (callable) - Optional - Callback when a participant disconnects.

### Request Example
```python
client.setup(
    service_type=ServiceType.INTERACTIVE_AVATAR,
    callbacks=HeyGenCallbacks(
        on_connected=my_connected_handler,
        on_participant_connected=my_participant_connected_handler
    )
)
```

### Response
None. This method configures the client instance.
```

--------------------------------

### Example Input Data for Word Timing Calculation

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.elevenlabs.tts.html

This is an example of the character alignment data structure expected by the calculate_word_times method. It includes characters, their start times, and end times.

```json
{
    "characters": [" ", "H", "e", "l", "l", "o", " ", "w", "o", "r", "l", "d"],
    "character_start_times_seconds": [0.0, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9],
    "character_end_times_seconds": [0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]
}

```

--------------------------------

### InterruptionFrame

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.frames.frames.html

Frame pushed to interrupt the pipeline. This frame is used to interrupt the pipeline. For example, when a user starts speaking to cancel any in-progress bot output. It can also be pushed by any processor.

```APIDOC
## InterruptionFrame

### Description
Frame pushed to interrupt the pipeline. This frame is used to interrupt the pipeline. For example, when a user starts speaking to cancel any in-progress bot output. It can also be pushed by any processor.
```

--------------------------------

### HeyGenClient.start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.client.html

Starts the client and establishes all necessary connections, including WebSocket and LiveKit, using the provided configuration and setting up audio processing.

```APIDOC
## start

### Description
Start the client and establish all necessary connections. Initializes WebSocket and LiveKit connections using the provided configuration. Sets up audio processing with the specified sample rates.

### Parameters
- **frame** (StartFrame) - Required - Initial configuration frame containing audio parameters
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.cartesia.tts.html

Starts the Cartesia HTTP TTS service, initializing it with the provided start frame.

```APIDOC
## async start(_frame : StartFrame_)

### Description
Start the Cartesia HTTP TTS service.

### Parameters
#### Path Parameters
* **frame** (StartFrame) - Required - The start frame containing initialization parameters.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.elevenlabs.tts.html

Starts the ElevenLabs HTTP TTS service. This method initializes the service with the provided start frame.

```APIDOC
## start

### Description
Start the ElevenLabs HTTP TTS service.

### Parameters
* **frame** (StartFrame_) - The start frame containing initialization parameters.
```

--------------------------------

### on_pipeline_started

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.observers.startup_timing_observer.html

Emits the startup timing report after the pipeline has fully started and all processors have processed the StartFrame.

```APIDOC
## on_pipeline_started

### Description
Emit the startup timing report when the pipeline has fully started.
Called by the `PipelineTask` after the `StartFrame` has been processed by all processors, including nested `ParallelPipeline` branches.

### Method
async
```

--------------------------------

### Start Neuphonic TTS Service

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.neuphonic.tts.html

Initiate the Neuphonic TTS service by calling the start method with the appropriate start frame.

```python
await service.start(StartFrame(...))
```

--------------------------------

### LiveAvatarApi.start_session

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.html

Starts a live avatar session. This method initiates the avatar's presence and prepares it for interaction.

```APIDOC
## LiveAvatarApi.start_session

### Description
Starts a live avatar session. This method initiates the avatar's presence and prepares it for interaction.

### Method
`LiveAvatarApi.start_session(session_id: str)`

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **session_id** (str) - Required - The ID of the live avatar session to start.

### Request Example
```json
{
  "session_id": "live_sess_abc123"
}
```

### Response
#### Success Response (200)
- **code** (int) - The status code of the response.
- **data** (LiveAvatarSessionData) - Data related to the started session.
  - **session_id** (str) - The ID of the session.
  - **livekit_url** (str) - The LiveKit URL for the session.
  - **livekit_client_token** (str) - The LiveKit client token.
  - **livekit_agent_token** (str) - The LiveKit agent token.
  - **max_session_duration** (int) - The maximum duration of the session in seconds.
  - **ws_url** (str) - The WebSocket URL for the session.
- **message** (str) - A message describing the result of the operation.

#### Response Example
```json
{
  "code": 200,
  "data": {
    "session_id": "live_sess_abc123",
    "livekit_url": "wss://livekit.example.com",
    "livekit_client_token": "lkct_ghi789",
    "livekit_agent_token": "lkat_jkl012",
    "max_session_duration": 3600,
    "ws_url": "wss://ws.heygen.com/live/live_sess_abc123"
  },
  "message": "Live avatar session started successfully."
}
```
```

--------------------------------

### GradiumTTSService.setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.gradium.tts.html

Sets up the processor with the required components and configuration.

```APIDOC
## setup

### Description
Set up the processor with required components.

### Method
async setup(_setup : FrameProcessorSetup_)

### Parameters
* **setup** (FrameProcessorSetup_) - Required - Configuration object containing setup parameters. [*]: Keyword-only parameters separator (PEP 3102)
```

--------------------------------

### HeyGenClient.setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.client.html

Sets up the client and initializes the conversation by establishing a new session with HeyGen’s API if one doesn’t already exist.

```APIDOC
## setup

### Description
Setup the client and initialize the conversation. Establishes a new session with HeyGen’s API if one doesn’t exist.

### Parameters
- **setup** (FrameProcessorSetup) - Required - The frame processor setup configuration.
```

--------------------------------

### Start TaskObserver

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.pipeline.task_observer.html

Starts all the asynchronous tasks associated with the managed proxy observers. This should be called after observers are added and before the pipeline starts processing frames.

```python
await task_observer.start()
```

--------------------------------

### HeyGenVideoService Setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.heygen.video.html

Sets up the HeyGen video service with necessary configuration, including initializing the HeyGen client and preparing for audio/video processing.

```APIDOC
## setup

### Description
Set up the HeyGen video service with necessary configuration. Initializes the HeyGen client, establishes connections, and prepares the service for audio/video processing. This includes setting up audio/video streams, configuring callbacks, and initializing the resampler.

### Parameters
* **setup** (FrameProcessorSetup_) - Configuration parameters for the frame processor.
```

--------------------------------

### Initializing ServiceSwitcher with Manual Strategy

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.pipeline.service_switcher.html

Demonstrates how to initialize a ServiceSwitcher with a list of services and specify the ServiceSwitcherStrategyManual for manual control.

```python
stt_switcher = ServiceSwitcher(
    services=[stt_1, stt_2],
    strategy_type=ServiceSwitcherStrategyManual
)
```

--------------------------------

### AsyncAITTSService.start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.asyncai.tts.html

Starts the Async TTS service with the provided start frame.

```APIDOC
## AsyncAITTSService.start

### Description
Start the Async TTS service.

### Parameters
- **frame** (StartFrame_) - Required - The start frame containing initialization parameters.
```

--------------------------------

### start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.elevenlabs.tts.html

Starts the ElevenLabs TTS service, initiating the speech synthesis process.

```APIDOC
## async start(_frame : StartFrame_) ##

### Description
Starts the ElevenLabs TTS service, initiating the speech synthesis process.

### Parameters
* **frame** (StartFrame) - Required - The start frame containing initialization parameters.
```

--------------------------------

### Start SoundfileMixer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.audio.mixers.soundfile_mixer.html

Initializes the mixer and loads all specified sound files. Requires the sample rate of the output transport.

```python
await mixer.start(sample_rate=44100)
```

--------------------------------

### MCPClient Start and Close

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.mcp_service.html

Manually starts and closes the persistent connection to the MCP server.

```APIDOC
## MCPClient Start and Close

### Description
Manually starts a persistent connection to the MCP server, opening the transport and initializing the session. The `close()` method can be called to terminate the connection.

### Methods

*   `async start() -> None`
    Starts a persistent connection to the MCP server.

*   `async close() -> None`
    Closes the persistent MCP connection. Safe to call multiple times or without having called start().
```

--------------------------------

### Install Pipecat Runner

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.run.html

Install the Pipecat runner package with the necessary dependencies for development.

```bash
pip install pipecat-ai[runner]

```

--------------------------------

### Handle Pipeline Started Event

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.observers.base_observer.html

Called when the pipeline has fully started, after the StartFrame has been processed by all processors.

```python
async on_pipeline_started()
```

--------------------------------

### GroqTTSService.setup

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.groq.tts.html

Sets up the Groq TTS service processor with the necessary components for operation.

```APIDOC
## async GroqTTSService.setup

### Description
Sets up the processor with required components. This method configures the service for operation.

### Parameters
* **setup** (FrameProcessorSetup) - Required - Configuration object containing setup parameters.
```

--------------------------------

### Start LMNT TTS Service

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.lmnt.tts.html

Starts the LMNT TTS service, accepting initialization parameters via a StartFrame.

```python
async def start(self, _frame: StartFrame_):
    """Start the LMNT TTS service.

    Parameters:
        frame – The start frame containing initialization parameters.
    """
    pass
```

--------------------------------

### Initializing ServiceSwitcher with Failover Strategy

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.pipeline.service_switcher.html

Shows how to set up a ServiceSwitcher with the ServiceSwitcherStrategyFailover. This strategy automatically switches to a backup service upon failure.

```python
switcher = ServiceSwitcher(
    services=[primary_stt, backup_stt],
    strategy_type=ServiceSwitcherStrategyFailover,
)

@switcher.strategy.event_handler("on_service_switched")
async def on_switched(strategy, service):
    # App decides when/how to recover the failed service
    ...
```

--------------------------------

### main()

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.html

The main entry point for the runner.

```APIDOC
## main()

### Description
Main function to start and run the Pipecat runner.

### Method
(Not specified, likely a function call)

### Parameters
(Not specified)
```

--------------------------------

### Azure HTTP TTS Service - start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.azure.tts.html

Starts the Azure HTTP TTS service and initializes the speech synthesizer.

```APIDOC
## start AzureHttpTTSService

### Description
Start the Azure HTTP TTS service and initialize speech synthesizer.

### Parameters
* **frame** (StartFrame_) - Required - Start frame containing initialization parameters.
```

--------------------------------

### Azure Streaming TTS Service - start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.azure.tts.html

Starts the Azure streaming TTS service and initializes the speech synthesizer.

```APIDOC
## start AzureTTSService

### Description
Start the Azure TTS service and initialize speech synthesizer.

### Parameters
* **frame** (StartFrame_) - Required - Start frame containing initialization parameters.
```

--------------------------------

### __init__

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.google.llm.html

Initializes the Google LLM service with API key, model, and other optional configurations.

```APIDOC
## __init__

### Description
Initialize the Google LLM service.

### Parameters
* **api_key** (str) - Required - Google AI API key for authentication.
* **model** (str | None) - Optional - Model name to use. Deprecated since version 0.0.105: Use `settings=GoogleLLMService.Settings(model=...)` instead.
* **params** (InputParams | None) - Optional - Optional model parameters for inference. Deprecated since version 0.0.105: Use `settings=GoogleLLMService.Settings(...)` instead.
* **settings** (GoogleLLMSettings | None) - Optional - Runtime-updatable settings for this service. When both deprecated parameters and _settings_ are provided, _settings_ values take precedence.
* **system_instruction** (str | None) - Optional - System instruction/prompt for the model. Deprecated since version 0.0.105: Use `settings=GoogleLLMService.Settings(system_instruction=...)` instead.
* **tools** (list[dict[str, Any]] | None) - Optional - List of available tools/functions.
* **tool_config** (dict[str, Any] | None) - Optional - Configuration for tool usage.
* **http_options** (HttpOptions | None) - Optional - HTTP options for the client.
* **kwargs** - Additional arguments passed to parent class.
```

--------------------------------

### Start AWSTranscribeSTTService

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.aws.stt.html

Initializes the connection to AWS Transcribe when the service starts. This is typically called internally by the Pipecat framework.

```python
await service.start(StartFrame())
```

--------------------------------

### __init__

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.openai.base_llm.html

Initializes the BaseOpenAILLMService with configuration parameters for interacting with the OpenAI API.

```APIDOC
## __init__

### Description
Initialize the BaseOpenAILLMService.

### Parameters
* **model** (str | None) - The OpenAI model name to use (e.g., “gpt-4.1”, “gpt-4o”). Deprecated since version 0.0.105: Use `settings=BaseOpenAILLMService.Settings(model=...)` instead.
* **api_key** (str | None) - OpenAI API key. If None, uses environment variable.
* **base_url** (str | None) - Custom base URL for OpenAI API. If None, uses default.
* **organization** (str | None) - OpenAI organization ID.
* **project** (str | None) - OpenAI project ID.
* **default_headers** (Mapping[str, str] | None) - Additional HTTP headers to include in requests.
* **service_tier** (str | None) - Service tier to use (e.g., “auto”, “flex”, “priority”).
* **params** (InputParams | None) - Input parameters for model configuration and behavior. Deprecated since version 0.0.105: Use `settings=BaseOpenAILLMService.Settings(...)` instead.
* **settings** (OpenAILLMSettings | None) - Runtime-updatable settings. When provided alongside deprecated parameters, `settings` values take precedence.
* **retry_timeout_secs** (float | None) - Request timeout in seconds. Defaults to 5.0 seconds.
* **retry_on_timeout** (bool | None) - Whether to retry the request once if it times out.
* **kwargs** - Additional arguments passed to the parent LLMService.
```

--------------------------------

### UserStartedSpeakingFrame

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.frames.frames.html

Frame indicating that the user turn has started. Emitted when the user turn starts, which usually means that some transcriptions are already available.

```APIDOC
## UserStartedSpeakingFrame

### Description
Frame indicating that the user turn has started. Emitted when the user turn starts, which usually means that some transcriptions are already available.
```

--------------------------------

### Setting up VADController with Task Manager

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.audio.vad.vad_controller.html

Illustrates how to set up the VADController by associating it with a BaseTaskManager. This is necessary before the controller can start processing frames or managing tasks.

```python
await vad_controller.setup(task_manager)
```

--------------------------------

### KokoroTTSService.setup()

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.kokoro.html

Sets up the Kokoro TTS service. This method is likely used for initialization or configuration before other service operations can be performed.

```APIDOC
## KokoroTTSService.setup()

### Description
Initializes or configures the Kokoro TTS service. This method should be called before utilizing other functionalities of the service.

### Method
(Not specified, likely a method call within a Python SDK)

### Parameters
(Not specified in source)

### Request Example
(Not specified in source)

### Response
(Not specified in source)
```

--------------------------------

### Start Async TTS Service

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.asyncai.tts.html

Initiates the Async TTS service using a provided start frame containing initialization parameters.

```python
await service.start(start_frame)
```

--------------------------------

### main

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.run.html

Starts the Pipecat development runner, parsing command-line arguments and launching a FastAPI server that supports multiple transports.

```APIDOC
## pipecat.runner.run.main(_parser: ArgumentParser | None = None)

### Description
Initializes and starts the Pipecat development runner. This function parses command-line arguments to configure the server, which supports WebRTC, Daily, and telephony transports concurrently. Clients specify their desired transport via the `/start` endpoint. If a transport is specified via the `-t` flag, the server will only support that single transport for `/start` requests and will display transport-specific startup details.

The runner automatically discovers and executes the `bot(runner_args)` async function defined in the calling module.

### Command-line Arguments
- `--host`: Server host address (default: `localhost`).
- `--port`: Server port (default: `7860`).
- `-t` or `--transport`: Restricts the server to a single transport (e.g., `daily`, `webrtc`, `twilio`, `telnyx`, `plivo`, `exotel`) and sets it as the default for `/start`. If omitted, all transports are supported.
- `-x` or `--proxy`: Public proxy hostname for telephony webhooks.
- `-d` or `--direct`: Connects directly to a Daily room (automatically sets transport to `daily`).
- `-f` or `--folder`: Path to the downloads folder.
- `--dialin`: Enables Daily PSTN dial-in webhook handling.
- `--esp32`: Enables SDP munging for ESP32 compatibility (requires `--host` to be an IP address).
- `--whatsapp`: Ensures required WhatsApp environment variables are present.
- `-v` or `--verbose`: Increases logging verbosity.

### Parameters
- `_parser` (ArgumentParser | None): An optional pre-configured argument parser. If None, a default parser is created.
```

--------------------------------

### Single Transport Bot Example

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.run.html

An example of a bot implementing the `bot(runner_args)` async function for a single transport, using DailyTransport.

```python
async def bot(runner_args: RunnerArguments):
    transport = DailyTransport(
        runner_args.room_url,
        runner_args.token,
        "Bot",
        DailyParams(...)
    )
    # Your bot logic here
    await run_pipeline(transport)

if __name__ == "__main__":
    from pipecat.runner.run import main
    main()

```

--------------------------------

### Setup LMNT TTS Processor

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.lmnt.tts.html

Sets up the LMNT TTS processor with the necessary components and configuration parameters provided in the setup object.

```python
async def setup(self, _setup: FrameProcessorSetup_):
    """Set up the processor with required components.

    Parameters:
        setup – Configuration object containing setup parameters.
    """
    pass
```

--------------------------------

### Initialize LLMSwitcher

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.pipeline.llm_switcher.html

Instantiate an LLMSwitcher with a list of LLM services. The default strategy is manual switching.

```python
llm_switcher = LLMSwitcher(llms=[openai_llm, anthropic_llm])
```

--------------------------------

### Start the Inworld TTS Service

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.inworld.tts.html

Start the Inworld TTS service by providing a StartFrame object. This method initiates the service's operation.

```python
await tts_service.start(frame=StartFrame())
```

--------------------------------

### Starting and Closing MCPClient Connection Manually

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.mcp_service.html

Illustrates the manual process of starting and closing a persistent connection to the MCP server using MCPClient.

```python
mcp = MCPClient(server_params=...)
await mcp.start()
try:
    ...
finally:
    await mcp.close()
```

--------------------------------

### Forward Pipeline Started Signal

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.pipeline.task_observer.html

Forwards the pipeline started signal to all observers managed by the TaskObserver. This is called internally when the pipeline begins execution.

```python
await task_observer.on_pipeline_started()
```

--------------------------------

### GoogleTTSService Initialization and Usage

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.google.tts.html

Demonstrates how to initialize the GoogleTTSService with credentials and settings, and how to use the run_tts method to synthesize speech from text.

```APIDOC
## GoogleTTSService

### Description
Provides real-time text-to-speech synthesis using Google Cloud’s streaming API for low-latency applications. Optimized for Chirp 3 HD and Journey voices with continuous audio streaming capabilities.

**Note:** Requires Google Cloud credentials via service account JSON, file path, or default application credentials (GOOGLE_APPLICATION_CREDENTIALS env var). Only Chirp 3 HD and Journey voices are supported. Use GoogleHttpTTSService for other voices.

### Initialization Example
```python
tts = GoogleTTSService(
    credentials_path="/path/to/service-account.json",
    settings=GoogleTTSService.Settings(
        voice="en-US-Chirp3-HD-Charon",
        language=Language.EN_US,
    )
)
```

### Method: run_tts

#### Description
Generate streaming speech from text using Google’s streaming API.

#### Parameters
- **text** (str) - Required - The text to synthesize into speech.
- **context_id** (str) - Required - The context ID for tracking audio frames.

#### Yields
- **Frame** - Audio frames containing the synthesized speech as it’s generated.
```

--------------------------------

### FishAudioTTSService.start

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.fish.tts.html

Starts the Fish Audio TTS service. This method is used to initiate the TTS service with the provided start frame, which contains initialization parameters.

```APIDOC
## FishAudioTTSService.start

### Description
Start the Fish Audio TTS service.

### Method
`start`

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
* **frame** (StartFrame_) - Required - The start frame containing initialization parameters.

### Request Example
```json
{
  "frame": {
    "initialization_parameters": {
      "api_key": "YOUR_FISH_AUDIO_API_KEY",
      "model_id": "s1"
    }
  }
}
```

### Response
#### Success Response (200)
None

#### Response Example
None
```

--------------------------------

### GroqLLMService.__init__

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.services.groq.llm.html

Initializes the Groq LLM service with API key and optional base URL, model, and settings.

```APIDOC
## GroqLLMService.__init__

### Description
Initialize Groq LLM service.

### Method Signature
`__init__(_*_ , _api_key : str_, _base_url : str = 'https://api.groq.com/openai/v1'_, _model : str | None = None_, _settings : GroqLLMSettings | None = None_, _** kwargs_)

### Parameters
* **api_key** (`str`) - The API key for accessing Groq’s API.
* **base_url** (`str`, optional) - The base URL for Groq API. Defaults to “https://api.groq.com/openai/v1”.
* **model** (`str | None`, optional) - The model identifier to use. Defaults to “llama-3.3-70b-versatile”. Deprecated since version 0.0.105: Use `settings=GroqLLMService.Settings(model=...)` instead.
* **settings** (`GroqLLMSettings | None`, optional) - Runtime-updatable settings. When provided alongside deprecated parameters, `settings` values take precedence.
* **kwargs** - Additional keyword arguments passed to OpenAILLMService.
```

--------------------------------

### Initialize Startup Timing Observer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.observers.startup_timing_observer.html

Instantiate the Startup Timing Observer, optionally specifying processor types to measure. If no types are provided, all non-internal processors will be measured.

```python
observer = StartupTimingObserver()
task = PipelineTask(pipeline, observers=[observer])
```

--------------------------------

### Building a Started Async Tool Message

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.processors.aggregators.async_tool_messages.html

Use this function to create a 'started' message when an asynchronous tool call begins. This informs the LLM that a task is in progress and results will follow.

```python
from pipecat.processors.aggregators.async_tool_messages import build_started_message

# Example usage:
# tool_call_id = "call_123"
# started_message = build_started_message(tool_call_id)
# llm_context.add_message(started_message)
```

--------------------------------

### Setup TwilioFrameSerializer

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.serializers.twilio.html

Sets up the serializer with pipeline configuration using a StartFrame. This method is part of the serializer's lifecycle.

```python
await serializer.setup(frame)
```

--------------------------------

### runner_port()

Source: https://reference-server.pipecat.ai/en/latest/api/pipecat.runner.html

Gets the port for the runner.

```APIDOC
## runner_port()

### Description
Retrieves the port number for the runner.

### Method
(Not specified, likely a function call)

### Parameters
(Not specified)
```
Tokens:
Raw