Polly Generative engine upgraded with more natural speech, 14 new voices, and 2 new languages

Amazon Polly’s Generative engine has received a model upgrade that improves speech naturalness and accuracy while maintaining low latency. The update introduces 14 new voices spanning nine locales, including Hindi and Finnish, which are now generally available for the first time. New additions include Kenneth and Nate for US English, Russell and Curtis for Australian English, Kiara and Arjun for Hindi and Indian English, Tina for German, Carlos for US Spanish, Leo and Carolina for Brazilian Portuguese, Hyunduk for Korean, and Suvi for Finnish.

The enhanced voice portfolio is part of a broader effort to make speech output more reliable, particularly for complex content such as numbers, email addresses, and street addresses. Users can continue leveraging Speech Synthesis Markup Language tags to fine-tune speech output as before. The low-latency streaming API supports real-time applications like contact centers, chatbots, and agentic voice systems built on large language models.

The upgrade expands Amazon Polly Generative to AWS GovCloud, broadening availability for public sector and regulated industries. All new Generative voices are now accessible in 11 AWS Regions, including US East, US West, multiple European and Asia Pacific locations, Canada, and AWS GovCloud (US-West).

Amazon Polly remains a fully managed text-to-speech service, and the model upgrade is designed as a drop-in replacement to improve existing implementations. Users can preview the new voices on the Amazon Polly Features page and consult the documentation and pricing page for further details on capabilities and costs.