Speech Synthesis Markup Language, or SSML, is a standardized markup language that provides a way to mark up text for changing how speech is synthesized. Numerous SSML tags are already supported by the Alexa Skills Kit, including: audio, break, p, phoneme, s, say-as, speak, and w.
Alexa now understands SSML Speechcons, which are special words and phrases that are pronounced more expressively by Alexa. Speechcons can be used in English (US) skills by adding a <say-as interpret-as="interjection"> tag around the speechcon you would like to use. Support for English (UK) and German skills coming soon.
For example, if you wanted Alexa to add emphasis when she says “bazinga,” you could create the following SSML:
Sometimes when I look at the Alexa skills you have all taught me, I just have to say,
In the above example, because "bazinga" appears inside the say-as tag, Alexa recognizes it as a speechcon and adds emphasis automatically. Here is an audio sample of how the emphasis sounds: Bazinga. As a good practice, you should also surround each speechcon with a pause.
There are over 100 speechcons that will be automatically recognized, here are just a few:
Abracadabra, achoo, aha, ahem, ahoy, all righty, aloha, baa, bada bing bada boom, bah humbug, bam, bang, batter up, bazinga, checkmate, cheerio, cheers, cheer up, and more!
You can check out the full list of supported Speechcons here.
For more information about getting started with Alexa and SSML, check out the following:
Speech Synthesis Markup Language (SSML) Reference
Alexa Skills Kit (ASK)
Alexa Dev Chat Podcast
Alexa Developer Forums