TTS Generation
Text-to-speech is more than reading words aloud. Shape how your content sounds — the tone, pacing, emotion, and personality — through intentional prompting.
Techniques
Direct the Tone
Specify emotional quality: 'warm and reassuring,' 'crisp and authoritative,' 'playful with a hint of mischief.' The voice should match the message's intent.
Control Pacing and Emphasis
Use markers for pauses, stress, and rhythm: 'Pause briefly after the question,' 'Emphasize the word never,' 'Read at a slow, deliberate pace.'
Set the Acoustic Context
Describe the imagined setting: 'as if narrating a nature documentary,' 'like a late-night radio host,' 'with the energy of a sports commentator.'
Handle Pronunciation and Style
Specify how to handle technical terms, acronyms, or names. Indicate dialect, formality level, and whether to use contractions or full forms.
Before & After
"Read this paragraph aloud."
"Narrate the following as a calm, thoughtful podcast host speaking to a curious audience. Pace: unhurried, with natural pauses after key insights. Emphasize words in bold. Pronounce 'CRISPR' as 'krisper.' Tone: intellectual but warm, like explaining something fascinating to a friend over coffee."
Explore TTS Prompts
Discover meta-prompts for crafting expressive text-to-speech outputs.
Visit the Library