Describe a voice.
Hear it say the line.
No audio to upload. Write the voice in ordinary words, in any language. It speaks one of ten languages. Listen, ask for a change, and save it with your other voices.
- No recording
- Any language
- 10 spoken languages
- Ask for a change
- Use it on a video
Free to start. Three creations on the free plan.
A clone needs a recording of that voice.
A description does not.
How it works
-
1. Describe the voice
Any language. Two words, or a paragraph. Age, accent, pace, mood, and what the voice is for.
-
2. Pick the language and the line
One of ten languages. The line is one sentence in that language, up to 180 characters. Saving keeps this clip and the words it spoke.
-
3. Listen, then change it
Warmer, faster, more energy, or your own note. The rest stays. Try again makes a different voice. Save keeps this one.
-
4. Name it and save it
It shows up with your created voices. Use it in Narrate, or in Text to Speech.
The screen
These are examples of what you type. Creating the voice happens after you sign in.
Describe the voice
Write it in any language, as short or as detailed as you like.
Language
The language this voice speaks.
Line to speak
One short sentence. Saving keeps this clip and the words it spoke.
Need changes
Write the change in any language. Everything else about the voice stays the same.
Which one you need
You already have the voice
Voice Cloning. Upload 10 to 15 seconds, or record live. Then type anything in that voice.
Go to Voice CloningYou can describe it
Voice Creator. No sample. Write the voice, hear the line, keep the one you want.
What you can make
The description can name the job. The saved voice is yours to reuse.
-
Onboarding and product narration
A calm voice for the first-run tour, or the same narrator on every demo.
-
Stories and audiobooks
A storyteller, slow and low, or a warmer reader for a long chapter.
-
Commentary
A faster, more energetic voice when the picture is a match or a launch.
-
Teaching and news
Clear, steady speech for a lesson or a short briefing.
-
A role, not a person
A narrator, a guide, a storyteller. Describe the voice. Do not name a real person to copy.
-
While you narrate a video
You can also describe a voice inside Video Narration. It speaks the narration language, when that language is one of the ten.
Honest details
-
Description and line are different
The description can be in any language. The spoken line has to be in the language you picked, one sentence, up to 180 characters.
-
Ten languages, and their accents
English, Spanish, Portuguese, French, German, Italian, Russian, Chinese, Japanese, Korean. Regional accents of those languages work, such as British, Indian English, Castilian, and Brazilian Portuguese. An accent from outside them is not supported.
-
English with another accent
If the description asks for a French accent while English is selected, you are asked to speak French instead. Save the voice in French, then use it in Text to Speech to speak English with that accent.
-
Not a copy of a real person
Describe the voice itself. Requests to copy a real person are not supported.
-
Each listen counts
A preview uses its length in seconds. A change, or Try again, is another creation. The free plan includes 3. Saving the voice is part of a paid plan.
-
Saved once, used again
After you save it, the voice is listed with your created voices. Use it in Narrate and in Text to Speech without describing it again.
Questions
The voice does not have to be recorded.
Describe it.
Ten languages. A change when you want one. Then it works like any saved voice.