Amazon Polly

Amazon Web Services' Polly TTS engine includes several models to accommodate different use cases.

Models

SignalWire supports the following three Amazon models.

Standard

Polly Standard is a traditional, cost-effective, and reliable TTS model.

Example voice ID string: amazon.Emma:en-GB or amazon.Emma:standard:en-GB

Neural

Polly Neural produces more natural, human-like speech than Polly Standard.

Example voice ID string: amazon.Kendra:neural:en-US

Generative

Polly Generative is Amazon's largest and most realistic model.

Example voice ID string: amazon.Danielle:generative:en-US

Languages

Consult AWS documentation for a comprehensive and up-to-date list of supported voices, as well as information on accented and fully bilingual voices.

List of supported voices

Most Amazon Polly voices support a single language. Select voices from this list, which includes Standard, Neural, and Generative models.

Bilingual pronunciation

All Amazon Polly voices support accented bilingual pronunciation through the use of the SSML lang tag.

Fully bilingual voices

Learn more about fully bilingual voices like Aditi, Kajal, Hala, and Zayd, which are designed to fluently speak two languages.

Usage

Amazon voice IDs conform to the following format:

amazon.<voice>:<model>:<language>

Parameter	Possible values	Description
`engine` required	`amazon` `polly` (will be deprecated in the future)	TTS engine
`voice` required	Choose from the Name/ID column	Voice ID
`model` optional	`standard`, `neural`, or `generative`	Amazon Polly model. Default: `standard`
`language` optional	Choose from the Language code column	Sets model language. Default: `en-US`

Examples:

amazon.Joanna
amazon.Joanna:standard:en-US
amazon.Kendra:neural:en-US
amazon.Matthew:neural:en-US
amazon.Danielle:generative:en-US
amazon.Amy:generative:en-GB
amazon.Aditi:standard:hi-IN

Simplified Format:

For voices that only support one model, you can omit the model parameter:

amazon.<voice>:<language>

Example: amazon.Aditi:hi-IN (Standard model implied)

note

The polly engine code is being deprecated. Use amazon instead.

Character limits

Amazon Polly has a limit of 3000 chargeable characters in a single request. If your TTS request is longer than 3000 characters, you will experience silence.

Examples

See how to use Amazon Polly voices on the SignalWire platform.

SWML
RELAY Realtime SDK
Call Flow Builder
cXML

Use the languages SWML method to set one or more voices for an AI agent.

version: 1.0.0
sections:
  main:
  - ai:
      prompt:
        text: Have an open-ended conversation about flowers.
      languages:
        - name: English
          code: en-US
          voice: amazon.Ruth:neural

Alternatively, use the say_voice parameter of the play SWML method to select a voice for basic TTS.

version: 1.0.0
sections:
  main:
  - set:
      say_voice: "amazon.Ruth:neural"
  - play: "say:Greetings. This is the Ruth voice from Amazon Polly's Neural text-to-speech model."

// This example uses the Node.js SDK for SignalWire's RELAY Realtime API.
const playback = await call.playTTS({ 
    text: "Greetings. This is the Ruth voice from Amazon Polly's Neural text-to-speech model.",
    voice: "amazon.Ruth:neural",
});
await playback.ended();

<?xml version="1.0" encoding="UTF-8"?>
<Response>
<Say voice="amazon.Ruth:neural">
    Greetings. This is the Ruth voice from Amazon Polly's Neural text-to-speech model.
</Say>
</Response>

Models​