For an audiobook, choose a voice with steady but not flat intonation, a moderate-to-slow pace and a mid-to-low pitch, so it stays comfortable over hours of listening. Two strong starting points are Michelle, whose intonation spans 7.3 semitones, and Michael, a male Studio voice at 115.5 Hz.
A narrator reads for hours, so the main risk is monotony. Intonation range shows whether a voice moves enough to hold a listener, and it should not move so much that it sounds like a performance. Michelle's 7.3 semitones is steady, narrower than about two-thirds of the female English voices.
Pace matters over a full book. Luna and Wayne read at 150 words per minute, slower than about 93 percent of English voices. That gives listeners room, but it can drag over a long chapter. Michelle at 167 words per minute and Christopher at 164 sit nearer the middle, a comfortable pace for long sessions.
Writing an audiobook script for an AI voice
An audiobook script is read as a book, so consistent structure matters more than it does in a short video.
Put each chapter title on its own line, spelled out the way you want it read, such as 'Chapter Seven: The Harbour'.
Tag each speaker consistently, such as 'Narrator' or 'MARA:', so every edit is easy to check.
Use full stops for pauses between paragraphs, and commas inside long sentences. Punctuation controls pauses.
Vary sentence length. A run of very short sentences sounds like a list.
Write numbers, dates and foreign names the way you want them read, and preview the first page of each new name.
Split long chapters into scenes, and preview one scene before you generate the whole chapter.
Male or female voice for an audiobook?
Both genders fit. Among the male picks, Eric's 106 Hz is near the bottom of the 36 male English voices, while Michael's 115.5 Hz is in the middle. Among the female picks, Michelle's 184 Hz sits in the lower-middle of the 39 female English voices, and Aria's 205 Hz is higher than 84 percent of them.
Choose the register that suits the book's narrator, and set the accent to the story's setting. Libby is British, so she suits a British setting, while Michelle and Michael are American.
Gerard, a Belgian male voice at 120 Hz, is higher-pitched than two-thirds of the seven male French voices.
Frequently asked questions
Look for steady intonation, a moderate-to-slow pace and mid-to-low pitch. Michelle (7.3-semitone range, 167 words per minute) and Michael (115.5 Hz) fit those measures, and Libby is the British female option at 151 words per minute.
Luna and Wayne, both Singaporean English voices, read at 150 words per minute, slower than about 93 percent of English voices in the library.
Libby is a British female voice with a median pitch of 196 Hz and an intonation range of 10.5 semitones, so she fits a British setting.
Eric, an American male voice with a median pitch of 106 Hz, sits in the lower part of the male group. Federico, a Nicaraguan male Spanish voice at 101 Hz, is the lowest-pitched male Spanish voice in the list.
Put each chapter title on its own line, tag speakers consistently, and use full stops and commas for pauses. Preview one scene before you generate the chapter.
The choice depends on the book's register and setting. Eric (106 Hz) and Michael (115.5 Hz) are the male picks, while Michelle (184 Hz) and Libby (196 Hz) are the female picks in the middle range.
Try these voices on your own script
Paste your text in Voice Studio, pick a voice from this page and preview it before you generate.