Choosing a voice

AI voices for audiobooks

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.

Recommended voices

Michelle

Female · American

GLOBAL

Steady female American narrator

Intonation range of 7.3 semitones, narrower than about two-thirds of female English voices, at 167 words per minute, which keeps long chapters even.

Pitch
184.2 Hz
Range
7.3 st
Pace
167 wpm
Use voice
Michael

Male · American

STUDIO

Studio male American narrator

Median pitch of 115.5 Hz, in the middle of the male group, at 171 words per minute with 7.6 semitones of intonation.

Pitch
115.5 Hz
Range
7.6 st
Pace
171 wpm
Use voice
Libby

Female · British

GLOBAL

British female narrator

Median pitch of 196 Hz and 10.5 semitones of intonation, at 151 words per minute, slower than nearly 90 percent of English voices.

Pitch
196.1 Hz
Range
10.5 st
Pace
151 wpm
Use voice
Christopher

Male · American

GLOBAL

Mid-range male American narrator

Median pitch of 110.5 Hz, lower than 77 percent of male voices, with 10.6 semitones of intonation at 164 words per minute.

Pitch
110.5 Hz
Range
10.6 st
Pace
164 wpm
Use voice
Eric

Male · American

GLOBAL

Low-pitched male American narrator

Median pitch of 106 Hz, lower than 89 percent of male English voices, and 11.0 semitones of intonation at 161 words per minute.

Pitch
106.4 Hz
Range
11.0 st
Pace
161 wpm
Use voice
Wayne

Male · Singaporean

GLOBAL

Slow male Singaporean narrator

Median pitch of 120 Hz with 8.5 semitones of intonation, at 150 words per minute, slower than about 93 percent of English voices.

Pitch
120.4 Hz
Range
8.5 st
Pace
150 wpm
Use voice
Luna

Female · Singaporean

GLOBAL

Slow female Singaporean narrator

Median pitch of 186 Hz and 7.1 semitones of intonation, at 150 words per minute, the same slow pace as Wayne.

Pitch
186.2 Hz
Range
7.1 st
Pace
150 wpm
Use voice
Aria

Female · American

GLOBAL

Higher-pitched female American narrator

Median pitch of 205 Hz, higher than 84 percent of female English voices, with 9.6 semitones of intonation at 161 words per minute.

Pitch
204.8 Hz
Range
9.6 st
Pace
161 wpm
Use voice

What makes a voice work for audiobooks

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.

Same criteria in other languages

Hindi

Beta · Female · 180.5 Hz

Beta, a female Studio voice at 181 Hz, is the lowest-pitched of the three female Hindi voices.

Urdu

Gul · Female · 189.3 Hz

Gul, a female Indian voice with a 5.4-semitone range, has the narrowest intonation of the two female Urdu voices, which keeps delivery steady.

Bengali

Nabanita · Female · 235.9 Hz

Nabanita, a female Bangladeshi voice at 236 Hz, speaks Bangladeshi Bangla, which suits storytelling and educational narration for Bangladesh.

Arabic

Shakir · Male · 110.7 Hz

Shakir, an Egyptian male voice at 111 Hz with an 11.4-semitone range, speaks Egyptian Arabic, a variety widely understood across the Arab world.

Spanish

Federico · Male · 100.7 Hz

Federico, a Nicaraguan male voice at 101 Hz, is the lowest-pitched of the 24 male Spanish voices.

Persian

Dilara · Female · 183.1 Hz

Dilara, a female Iranian voice at 183 Hz, suits audiobooks in standard Iranian Persian.

Tamil

Kumar · Male · 145.6 Hz

Kumar, a Sri Lankan male voice at 146 Hz, suits listeners who want a local-sounding Sri Lankan Tamil narrator.

Portuguese

Thalita · Female · 202.6 Hz

Thalita, a Brazilian female voice at 203 Hz, sits in the upper-middle of the four female Portuguese voices and is multilingual.

French

Gerard · Male · 120.3 Hz

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.

Open Voice Studio