Amazon Polly text to speech¶

Status: Current | Last reviewed: 2026-09-25

Add a Text-to-Speech stage, choose Amazon as the provider and Polly as the service, then use the automatically selected default Integration Profile or choose another. In Configuration → Integrations, enable the Text to Speech capability on the Amazon profile you want to use. Only enabled platform integrations and integrations in your current tenant are offered.

Choose the AWS Region, Language, and Voice. Voices load from Amazon Polly using the selected integration and region. Choosing a language narrows the voice list; choosing a voice keeps its language and synthesis engine compatible. Engine offers the engines supported by the selected voice. Use Play sample to hear the same sentence used by the other voice previews: “The sky is the limit when the business reaches for the stars.” Audio rates are handled automatically by the channel and provider adapter.

The same form appears under Properties → Auxiliary Services → Private prompt voice (TTS) when the fallback service is Polly. Default selected means the agent still inherits its default. Change a setting to override it for this agent, or choose Use defaults to remove the override.

The private voice initially follows the main agent where possible. You can choose a different voice; the change applies only to this agent.

A voice may support different engines in different regions. If the requested engine is unavailable, the runtime uses an available engine. If voices fail to load, check the integration and region, then choose Try again. Published agents must be edited through a draft.