Choosing a voice

AI voices for podcasts

For podcasts, choose a voice with a moderate intonation range and a moderate pace, so the delivery stays conversational from the first minute to the last. Jenny reads at 170 words per minute with a 10.0-semitone range, and Andrew has a 12.7-semitone range with a 115.3 Hz median pitch.

Recommended voices

Jenny

Female · American

GLOBAL

Friendly female American host

Pace of 170 words per minute and a 10.0-semitone range, wider than about 66% of female English voices.

Pitch
181.3 Hz
Range
10.0 st
Pace
170 wpm
Use voice
Ava

Female · American

GLOBAL

American female narrator

Intonation range of 10.6 semitones, wider than about 76% of female English voices, with a 200.5 Hz median pitch, higher than about 74% of them.

Pitch
200.5 Hz
Range
10.6 st
Pace
179 wpm
Use voice
Sonia

Female · British

GLOBAL

British female narrator

Intonation range of 11.2 semitones, wider than about 79% of female English voices, at 171 words per minute.

Pitch
193.1 Hz
Range
11.2 st
Pace
171 wpm
Use voice
Andrew

Male · American

GLOBAL

Multilingual American male host

Intonation range of 12.7 semitones, wider than about 71% of male English voices, with a 115.3 Hz median pitch. The voice can also speak Spanish, French, German and Portuguese.

Pitch
115.3 Hz
Range
12.7 st
Pace
181 wpm
Use voice
Thomas

Male · British

GLOBAL

British male narrator

Median pitch of 119.2 Hz and an 11.2-semitone range, wider than about 66% of male English voices, at 168 words per minute.

Pitch
119.2 Hz
Range
11.2 st
Pace
168 wpm
Use voice
Christopher

Male · American

GLOBAL

American male narrator

Intonation range of 10.6 semitones, wider than about 54% of male English voices, at 164 words per minute.

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

Male · British

STUDIO

British male Studio narrator

Median pitch of 84.7 Hz, the lowest of the 36 male English voices we measured, with an 11.0-semitone range at 180 words per minute.

Pitch
84.7 Hz
Range
11.0 st
Pace
180 wpm
Use voice

What makes a voice work for a podcast

A host talks for long stretches, so the voice should not work too hard. Moderate range and moderate pace are the two numbers that matter. A very narrow range can sound flat over a long episode, and a very wide one asks more of the listener.

The English picks sit in a tight band. Their intonation ranges run from 10.0 to 12.7 semitones, and their paces from 164 to 181 words per minute.

Pitch is a separate choice. Lewis's 84.7 Hz median pitch is the lowest of the male picks, while Ava's 200.5 Hz sits higher than about 74% of female English voices.

Writing a podcast script for an AI voice

Podcast scripts should read like speech. Write the way you talk, then read the script aloud once before generating anything.

  • Use short paragraphs with one idea each, and open each one with a clear first sentence.
  • Open the episode with a hook of one or two sentences that states what the listener will learn.
  • Write numbers in words, such as 'two hundred' rather than '200', and spell out acronyms the first time they appear.
  • Keep a list of names and their spellings, and test any unfamiliar name in the preview.
  • Mark a change of segment with a blank line and a short bridge sentence, such as 'Now, the second tip.'
  • Use commas and full stops for pauses, and an ellipsis only where you want a longer hesitation.

Male or female voice for a podcast?

The contrast is pitch. The male picks sit between 84.7 Hz (Lewis) and 119.2 Hz (Thomas). The female picks sit between 181.3 Hz (Jenny) and 200.5 Hz (Ava).

Intonation range is similar on both sides: 10.6 to 12.7 semitones for the male picks and 10.0 to 11.2 for the female picks. Pick the pitch that suits the host, and keep the range in the moderate band.

Same criteria in other languages

Hindi

Madhur · Male · 132.5 Hz

Madhur, a male Indian voice with a 132.5 Hz median pitch.

Urdu

Uzma · Female · 206.4 Hz

Uzma, a female Pakistani voice with a 206.4 Hz median pitch.

Bengali

Tanishaa · Female · 255.5 Hz

Tanishaa, a female Indian voice with a 12.8-semitone range.

Arabic

Noura · Female · 214.0 Hz

Noura, a female Kuwaiti voice with a 10.5-semitone range.

Spanish

Victor · Male · 122.4 Hz

Victor, a male Puerto Rican voice with a 7.4-semitone range.

Persian

Dilara · Female · 183.1 Hz

Dilara, a female Persian voice with a 9.1-semitone range.

Tamil

Valluvar · Male · 153.3 Hz

Valluvar, a male Indian voice with a 153.3 Hz median pitch.

Portuguese

Francisca · Female · 255.0 Hz

Francisca, a female Brazilian voice with a 255.0 Hz median pitch.

French

Thierry · Male · 158.4 Hz

Thierry, a male Canadian voice with an 11.9-semitone range.

Frequently asked questions

Start with a moderate range and pace. Jenny has a 10.0-semitone range at 170 words per minute, and Andrew has a 12.7-semitone range with a 115.3 Hz median pitch.
Ava and Sonia are the female picks with the wider ranges, at 10.6 and 11.2 semitones. Their median pitches are 200.5 Hz and 193.1 Hz. Preview both on your first two minutes.
Yes. Rewrite the article as spoken sentences first, with short paragraphs and names spelled out. Then choose a voice from the picks above and preview one section.
Madhur is a male Indian Hindi voice with a 132.5 Hz median pitch. Uzma is a female Pakistani Urdu voice with a 206.4 Hz median pitch.
Andrew can also speak Spanish, French, German and Portuguese in the same voice, so one male host can read an English episode and a Spanish version. Check the pronunciation of names in a preview.
Realism depends on the script as much as the voice. Our measurements cover pitch, range, pace and tone, so compare picks on those numbers and preview the same paragraph in each.

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