Artificial intelligence news.
Straight from the original announcements.

Amazon Polly makes voice cloning generally available to AWS customers

November 19, 2026, 20:53By the AI News Bites Editorial TeamSource: AWS
Amazon Polly makes voice cloning generally available to AWS customers
Image: AWS

Amazon Polly has made voice cloning generally available as a gated release, allowing customers to create a synthetic voice from a few minutes of recordings supplied by a consenting speaker. The cloned voice can then generate speech through Polly’s generative engine.

Key points
  • Amazon Polly voice cloning is generally available as a gated release.
  • The feature supports English, Spanish, German, French, Italian, Portuguese and Hindi.
  • Voice creation typically takes 5-10 minutes and requires speaker consent.
  • Cloned voices are encrypted, access-controlled and scoped to the creating AWS account.
  • Voice creation is free, while cloned voice usage is pay-as-you-go.

Custom voices in seven languages

Voice cloning launches with English, Spanish, German, French, Italian, Portuguese and Hindi. Each cloned voice speaks the language in which it was recorded, so multilingual experiences require approved voices created separately for each supported language.

AWS says the cloned voices offer similar naturalness and accuracy to its other generative voices. The company’s examples include a game character, a customer support line, lesson narration and brand audio. Potential uses include interactive voice response systems, voice agents, mobile apps, in-store announcements, advertising, connected devices, audiobooks, podcasts, courses and videos.

For businesses, the feature can provide a consistent voice across channels and allow teams to update prompts without booking another recording session. Content creators and publishers can also use custom voices for new chapters, updates and localized material.

How Amazon Polly voice cloning works

Customers record one speaker and submit between 1 and 10 recordings when calling the CreateVoice API. They must provide TermsAccepted=True to confirm that they have the speaker’s consent. Recordings can be sent through the API, uploaded in the console or stored in an Amazon Simple Storage Service bucket in the AWS Region where the voice will be created.

Amazon Polly returns a voice ID beginning with vc- and builds the voice asynchronously. Creation typically takes 5-10 minutes. Customers can call GetVoice until the voice becomes active or the process fails, then use the voice with SynthesizeSpeech for real-time audio or StartSpeechSynthesisTask for long-form audio written to S3. A voice can be removed with DeleteVoice.

AWS recommends recording two to ten minutes of natural speech in a quiet room, using one speaker and consistent microphone placement. Customers can submit up to 15 minutes of audio in total, with recordings of up to a minute each. AWS recommends mono WAV, 16-bit, 24 kHz recordings.

Consent, access and availability

Amazon Polly requires customers to confirm consent each time they create a voice. AWS says cloned voices are scoped to the account that created them, with access controlled through AWS Identity and Access Management. Voices are stored encrypted, and customers can use their own AWS Key Management Service key. Generated speech from a cloned voice also carries a watermark.

AWS says it does not retain a copy of the source recordings after creating the voice or use them to train Amazon Polly models. Deleting a voice removes the associated data immediately. The company recommends keeping records of consent, deleting a voice if consent is withdrawn and telling listeners when they are hearing synthetic speech where appropriate.

Voice cloning is available in all AWS Regions supported by the Amazon Polly generative engine, including the AWS GovCloud (US) Region. Access is already enabled for select Amazon Polly customers. Other customers can contact their AWS account team or apply for access. Building a cloned voice is free for Amazon Polly customers, while use of the cloned voice is charged on a pay-as-you-go basis.

Related news

AWS brings PolyAI’s Dialog-RSN-1 voice model to SageMaker JumpStart
ModelsNovember 19, 2026, 18:28

AWS brings PolyAI’s Dialog-RSN-1 voice model to SageMaker JumpStart

PolyAI’s Dialog-RSN-1 reasoning model for real-time customer conversations is now available through AWS Marketplace and Amazon SageMaker JumpStart. AWS customers can…

Stretto builds Tara, an AI voice agent for bankruptcy intake
Products and toolsNovember 18, 2026, 19:58

Stretto builds Tara, an AI voice agent for bankruptcy intake

Stretto has built Tara, an AI voice agent that conducts bankruptcy intake interviews by phone and sorts leads using attorney-defined rules. The system, developed with…

Amazon Polly upgrades Generative engine with 14 new voices, Hindi and Finnish
Products and toolsOctober 22, 2026, 20:06

Amazon Polly upgrades Generative engine with 14 new voices, Hindi and Finnish

AWS has upgraded the Amazon Polly Generative engine to produce more expressive and more accurate speech with low latency. The update adds 14 new voices across 9 locales…