Amazon Polly adds voice cloning in seven languages as a gated release
Building a cloned voice is free. Amazon Polly charges pay-as-you-go for usage, and generated audio carries a watermark.
What happened
Amazon Polly is AWS's text-to-speech service. It can now copy a person's voice from a few minutes of recordings. The copy can then read any text you give it.
Why it matters to you
Companies can use one consistent voice for phone lines, apps, audiobooks and games without booking a new studio session whenever a script changes.
The takeaway
Amazon now lets approved customers turn a consenting speaker's recordings into a reusable synthetic voice.
What it means for you
Treat voice creation as a provisioning step: call CreateVoice, poll GetVoice, then pass the voice ID to SynthesizeSpeech or StartSpeechSynthesisTask with Engine=generative.
You must confirm the speaker's consent for each voice you create, and AWS recommends keeping a consent record and telling listeners when a voice is synthetic.
AWS says building a cloned voice is free and usage is billed pay-as-you-go, with details on the Amazon Polly Pricing page.