AWS Polly

AWS Polly is a text-to-speech plugin. It synthesizes the agent’s text responses into natural-sounding speech using Amazon Polly’s neural, standard, generative, and long-form voice engines.

Setup

Set your AWS access key in the worker environment. Get your access keys from the AWS console:

export AWS_ACCESS_KEY_ID=<key>

Usage

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

Python

from zrt.plugins import AWSPollyTTS

tts = AWSPollyTTS()

# Pipeline(tts=tts, ...)

Parameters

Parameter	Type	Default	Description
`aws_access_key_id`	`str \| None`	`None`	AWS access key ID. Falls back to the `AWS_ACCESS_KEY_ID` environment variable when unset.
`aws_secret_access_key`	`str \| None`	`None`	AWS secret access key paired with `aws_access_key_id`. Falls back to the `AWS_SECRET_ACCESS_KEY` environment variable when unset.
`aws_session_token`	`str \| None`	`None`	Temporary session token for short-lived IAM role credentials (e.g. from `sts:AssumeRole`). Falls back to the `AWS_SESSION_TOKEN` environment variable when unset. Omit for long-term credentials.
`region`	`str`	`"us-east-1"`	AWS region hosting the Polly endpoint, e.g. `"us-east-1"`, `"eu-west-1"`, `"ap-southeast-1"`.
`voice`	`str`	`"Joanna"`	Polly voice ID to use for synthesis (US English, female, neural). Other popular options include `"Matthew"` (US English, male), `"Amy"` (British English, female), `"Brian"` (British English, male), and `"Celine"` (French, female). See the AWS Polly voice list for the full catalogue.
`engine`	`str`	`"neural"`	Polly synthesis engine. One of `"neural"` (high-quality neural TTS; supports a subset of voices), `"standard"` (concatenative synthesis; broadest voice support), `"generative"` (expressive speech; select voices only), or `"long-form"` (optimized for longer text passages; select voices only).
`speed`	`float`	`1.0`	Playback rate multiplier (`1.0` is normal speed).
`pitch`	`float`	`0.0`	Pitch adjustment in semitones (`0.0` is normal).

Synthesized audio is streamed back to the caller after the LLM produces a response.

Plugins

Inference Gateway

Speech-to-text

LLM

Text-to-speech

Realtime models

Avatars

Turn detection

Voice activity detection

Noise cancellation

Setup

Usage

Parameters

​Setup

​Usage

​Parameters

Setup

Usage

Parameters