Amazon Polly

Amazon Polly is AWS's text-to-speech service, turning written text into spoken audio through a simple API call. It supports dozens of voices across 30+ languages, including both standard and neural voices, and offers a Newscaster style plus a Brand Voice option for companies that want a custom synthetic voice built from their own recordings. Output can be tuned with SSML tags for pronunciation, pitch, speed, and pauses, and Polly can stream audio in real time or generate files in formats like MP3 and OGG.

Developers building IVR systems, e-learning platforms, audiobook and news-reading apps, and accessibility features use Polly because it plugs directly into other AWS services and scales with pay-per-character pricing rather than requiring voice actors or fixed licensing fees. It competes with Google Cloud Text-to-Speech, Microsoft Azure Neural TTS, and newer voice AI companies like ElevenLabs, and while its neural voices are solid and reliable, it's generally seen as less expressive and less state-of-the-art than the newest generation of AI voice generators. Its real strength is being deeply embedded AWS infrastructure: predictable pricing, broad language coverage, and easy integration for teams already running on AWS rather than being the most cutting-edge voice quality on the market.

Most people never interact with Polly directly. It works quietly behind the scenes in call centers, transit announcements, accessibility readers, and voice-enabled apps, which makes it more of a backend utility than a consumer-facing product.