SSML

<?xml version="1.0" encoding="UTF-8"?>
<speak version="1.1"
       xmlns="http://www.w3.org/2001/10/synthesis"
       xml:lang="en-US">
  <p>
    <s>Welcome to the <emphasis level="strong">Speech Synthesis Markup Language</emphasis> demo.</s>
    <s>SSML gives fine-grained control over text-to-speech output.</s>
  </p>
  <p>
    <s>Rate can be <prosody rate="slow">slow</prosody>,
       <prosody rate="medium">medium</prosody>, or
       <prosody rate="fast">fast</prosody>.</s>
  </p>
  <p>
    <s>Pitch can be <prosody pitch="low">low</prosody> or
       <prosody pitch="high">high</prosody>.</s>
  </p>
  <p>
    <s>A short pause <break time="300ms"/> can be inserted mid-sentence.</s>
  </p>
  <p>
    <s>Numbers: <say-as interpret-as="cardinal">42</say-as>,
       dates: <say-as interpret-as="date" format="mdy">01/01/2025</say-as>,
       and characters: <say-as interpret-as="characters">SSML</say-as>.</s>
  </p>
  <p>
    <s>Custom pronunciation: <phoneme alphabet="ipa" ph="es-es-em-el">SSML</phoneme>.</s>
  </p>
</speak>