Speed, pitch and rhythm: what makes speech sound natural

Updated October 31, 2026

Two voices can say exactly the same words and sound completely different. The difference is prosody: the speed, pitch, rhythm, pauses and stress of speech.

Speaking rate

Speaking rate is usually measured in words per minute (wpm). In English, relaxed conversation is often around 150 wpm, and audiobook narrators typically read somewhere around 150 to 160 wpm. Presenters and teachers tend to speak a little slower so listeners can follow.

As a rough guide for planning audio, 1,000 characters of English text, about 160 to 180 words, makes a little over a minute of speech at a normal pace.

When to change the speed:

• Slower (0.75× to 0.9×) for language learning, complex technical content, or listeners who prefer more time.
• Normal (1×) for most narration and listening.
• Faster (1.25× to 2×) for experienced listeners who want to get through a lot of material. Many screen reader users listen much faster than normal speech once they are used to it.

Pitch

Pitch is how high or low a voice sounds. Every voice has a natural pitch range, and pitch moves constantly while we speak: it rises in many questions, falls at the end of statements, and jumps on words we want to emphasize. This movement is called intonation.

Lowering the pitch a little can make a voice sound calmer or more serious; raising it can make it sound brighter. Big changes quickly sound unnatural, so small adjustments work best.

Pauses

Pauses give listeners time to take in what was said. Speakers pause at the end of sentences, between paragraphs, and before important points. In text to speech, punctuation is what creates pauses. Our Pause between sentences option lets you lengthen them for a slower, clearer delivery. See how to write text that sounds good read aloud.

Stress and rhythm

In every sentence, some words carry more weight than others. "I didn't say she took it" and "I didn't say she took it" mean different things. Languages also have different overall rhythms: English tends to squeeze unstressed syllables between stressed ones, while languages like Spanish and French give syllables a more even length. Neural voices learn these patterns from recordings, which is a big reason they sound more natural than older voices.

Emotion and style

We change all of these together to express feeling. Excitement tends to mean faster speech and higher, more varied pitch. Sadness tends to be slower, lower and flatter. Our Style options, such as Happy, Calm, Serious or Sad, adjust speed, pitch, volume and expressiveness together to give the voice a different mood.

Finding the right settings

1. Start with the defaults and a natural or premium voice.
2. Listen to a paragraph and adjust the speed first.
3. Then try a style that suits your content.
4. Only change the pitch if the voice still doesn't feel right.

Try it now: the options are under the text box.