Azure

Azure is a speech-to-text plugin. It transcribes the caller’s audio into text using Microsoft Azure’s Speech service for the LLM.

Setup

Set your Azure Speech key in the worker environment. Create a Speech resource and key in the Azure portal:

export AZURE_SPEECH_KEY=<key>

Import the plugin and pass it to the pipeline’s stt slot.

from zrt.plugins import AzureSTT

stt = AzureSTT()

# Pipeline(stt=stt, ...)

Parameter	Type	Default	Description
`speech_key`	`str`	`None`	Azure Speech resource subscription key. Falls back to the `AZURE_SPEECH_KEY` environment variable when unset.
`speech_region`	`str`	`None`	Azure region identifier for the Speech resource (e.g. `"eastus"`, `"westeurope"`, `"southeastasia"`). Falls back to the `AZURE_REGION` environment variable, then `"eastus"`. Must match the region where your Azure Speech resource was created.
`language`	`str`	`"en-US"`	BCP-47 locale tag for the spoken language. Over 100 locales are supported.
`sample_rate`	`int`	`16000`	PCM input sample rate in Hz sent to Azure.
`enable_phrase_list`	`bool`	`False`	When `True`, the phrases supplied in `phrase_list` are registered as recognition hints to improve accuracy for domain-specific terms.
`phrase_list`	`list[str]`	`None`	List of words or phrases to use as recognition hints when `enable_phrase_list` is `True`.

Transcribed text passes to the LLM once turn detection decides the caller has finished speaking.