WORLDTECHAmazon Polly adds 14 voices and Hindi and Finnish to Generative engine
Amazon Web Services has made a major upgrade to the Amazon Polly Generative text-to-speech engine generally available in production. AWS says the engine now produces more natural and more accurate speech with low latency, and the update adds 14 new voices across nine locales along with generative voices for Hindi and Finnish.
What the upgrade includes
AWS published the details in a post on its Machine Learning Blog, describing the updated engine as offering more expressive and more accurate speech output with low latency. Low latency, the delay between a request and the returned audio, is among the gains the company highlights.
The new voices cover US English (Kenneth and Nate), Australian English (Russell and Curtis), Indian English (Kiara and Arjun), German (Tina), US Spanish (Carlos), Brazilian Portuguese (Leo and Carolina), Korean (Hyunduk), Hindi (Kiara and Arjun) and Finnish (Suvi). Across the whole generative catalogue there are now 57 voices in 11 languages and 25 locales, and Hindi and Finnish are generally available on the engine for the first time.
Existing applications keep working
The change is a drop-in upgrade for existing generative voices and needs no code changes. Voice names and API calls stay the same, so applications already built on Polly continue to run as they do today.
According to AWS, the engine reflects emotion and prosody more closely to the input text without manual setup. AWS has also published sample audio so that users can compare the output for themselves.
What Polly is used for
Amazon Polly is a fully managed AWS service that turns written text into speech that sounds natural. AWS names contact centres, large language model-based agentic applications, media and entertainment, e-learning and accessibility among the areas that need voice output reliable at scale.