AI service
Amazon Polly
Converts text to lifelike speech.
Key points
- Converts text input into spoken audio.
- Supports multiple voices, languages, and speech styles depending on the selected voice.
- Can use SSML to control pronunciation, pauses, emphasis, and speech formatting.
- Produces audio output for applications, devices, and content workflows.
- Complements Transcribe, which performs the opposite direction from speech to text.
When to use it
- Choose Polly when an application needs to read text aloud.
- Use it for voice responses, accessibility, audio content generation, and call center prompts.
- Use SSML when pronunciation or speech pacing needs control.
Exam tips
- Polly is text-to-speech; Transcribe is speech-to-text.
- Lex builds conversational bots and can use speech interfaces, but Polly specifically generates speech audio.
- Translate changes written language and can be combined with Polly for multilingual audio.
- The input is text and the output is audio.