# Amazon Polly

> An AWS service that turns written text into spoken audio.

- Level: Intermediate
- Topics: AWS, AI
- URL: https://tidbit.legios.com.ar/amazon-polly/
- In Spanish: [Amazon Polly](https://tidbit.legios.com.ar/es/amazon-polly/)

**Amazon Polly** is text-to-speech as an API. You send it a sentence, pick a voice and a language, and it sends back an audio file (MP3, for example) or a stream of audio that sounds like a person reading the text aloud. You pay for the amount of text you convert, and you can keep and replay the audio you generated without paying again.

It offers several families of voices. The older **standard** voices sound clearly synthetic; the **neural**, **long-form** and **generative** voices use newer machine learning models and sound much closer to a real narrator. If you need control over how something is said, you can wrap the text in **SSML**, a small markup language that marks pauses, emphasis or how to pronounce an acronym.

Typical uses: reading news articles aloud in an app, voices for e-learning courses and games, phone menus, and accessibility features for people who cannot easily read a screen. It is the opposite direction of Amazon Transcribe, which turns speech into text.

## An analogy

A voice actor on call who never gets tired: you slip a script under the door, choose who should read it, and a recording comes back a moment later.

## Keep in mind

- Converts text into natural-sounding speech in many languages and voices.
- Standard voices are cheaper; neural and generative voices sound more human.
- SSML lets you control pauses, emphasis and pronunciation.
- You pay per amount of text converted; replaying saved audio is free.

## Go deeper

- [What is Amazon Polly? (official docs)](https://docs.aws.amazon.com/polly/latest/dg/what-is.html)
- [Amazon Polly product page](https://aws.amazon.com/polly/)
- [Speech synthesis on Wikipedia](https://en.wikipedia.org/wiki/Speech_synthesis)

## Related terms

- [Text-to-speech](https://tidbit.legios.com.ar/text-to-speech/): Technology that reads written text aloud in a synthetic voice.
- [API](https://tidbit.legios.com.ar/api/): The menu of things one program lets another program ask for.
- [Machine learning](https://tidbit.legios.com.ar/machine-learning/): Teaching computers by showing them examples instead of writing every rule.
- [Web accessibility](https://tidbit.legios.com.ar/web-accessibility/): Making websites and apps usable by everyone, including people with disabilities.
- [Amazon Transcribe](https://tidbit.legios.com.ar/amazon-transcribe/): An AWS service that listens to audio and writes down what was said.