AI character voices: covers and text-to-speech

What is an AI character voice generator?

An AI character voice generator takes audio or text and gives it back in the voice of a character. On Jammable that voice is a trained AI voice model: someone in the community trained it on recordings of one voice, and Jammable uses it to re-sing a vocal (a cover) or to read out words you type (text-to-speech).

Jammable has more than 60,000 community voices. Character voices are a big part of the library, and they are grouped into three categories, so you don't have to dig through music voices to find them.

Where to find character voices

Open the anime voices, cartoon voices or gaming voices directly, or start from the full voice library and tap a category.

How to find the right character voice

There are three ways in, and they work together:

• Search. Type the name of the character into the search box in the voice library or in the voice picker on the cover maker. The same character can have several versions made by different people, so it is worth comparing a few.
• Sort. The library sorts by Trending today, Most used, Most liked and Newest. Trending shows what people are using right now; Most used and Most liked are a good shortcut to voices that already sound right.
• Browse a category. The cover maker's voice picker has Anime, Cartoon and Gaming tabs next to Trending, Recently used and My voices, so you can switch voices without leaving the page.

Each voice has its own page, and most have demos to listen to. Listen before you pick: two versions of the same character can sound quite different.

How to make a cover in a character voice

1. 1

   Open the cover maker

   Go to Create, or press the button on a voice's page to start a cover with that voice already picked.

2. 2

   Pick the character voice

   Press Pick a voice, then search or open the Anime, Cartoon or Gaming tab. Add more voices with Add a voice to make a duet of up to 6 voices.

3. 3

   Add your audio

   Upload a file, record straight from your microphone, or choose AI song to describe a style and have a new song written for the voice to sing.

4. 4

   Set the sound

   Change the pitch in semitones if the character sings higher or lower than the original vocal, and add autotune or an effect if you want one. If your audio is only vocals, turn on My audio is just vocals.

5. 5

   Create the cover

   Press Create cover. Jammable separates the vocal, re-sings it in the character's voice and mixes it back with the music.

How to use character text-to-speech

1. 1

   Pick the character voice

   On Create, press Pick a voice and choose the character, the same way as for a cover.

2. 2

   Choose Type words

   Under the audio options, choose Type words and write what the character should say, up to 2,000 characters.

3. 3

   Pick a base voice, accent and delivery

   Choose a female or male base voice, one of 9 accents (American, British, French, German, Spanish, Italian, Korean, Australian or Japanese) and one of 7 deliveries: default, whispered, happy, sad, angry, shouting or terrified.

4. 4

   Create it

   Press Use these words, then create. Jammable reads your text in the base voice and converts it into the character's voice.

Cover or text-to-speech?

Which should I use?

• Best for A character singingCover

  Add a vocal and Jammable re-sings it in the character's voice, keeping the melody and timing of the original.

• Best for A character saying a lineText-to-speech

  Type the line, pick a delivery like happy or shouting, and you get a spoken clip without recording anything.

• Best for A character singing new lyricsAI song

  Describe a style, write your own lyrics or let AI write them, and the character voice sings the result.

• Best for Two characters in one trackDuet

  Add up to 6 voices to one cover and let Jammable switch between them, or choose who sings each part.

Tips for a character voice that sounds right

• Choose a voice trained on clean audio. A voice learns everything in its training audio, including music, crowd noise and echo. Listen to its demos: if it sounds muddy or has music bleeding through, try another version of the same character.
• Check how much it is used. Sort by Most used or Most liked. Voices that many people come back to usually sound good.
• Match the voice's range. A character voice works best near the pitch it was trained on. If a high-pitched cartoon voice sounds strained on a low vocal, raise the pitch; the cover maker suggests +12 semitones going from a male to a female voice, and -12 the other way.
• Give it a clean vocal. A single, clear lead vocal gives the cleanest cover. If the song has harmonies or ad-libs, turn on Split out backing vocals.
• Go easy on effects. Character voices often have a strong tone of their own. Heavy autotune or reverb on top can hide it.
• For text-to-speech, write how the character talks. Short sentences and punctuation give clearer phrasing, and the delivery setting does more for emotion than extra words.

How character voices get made

Every voice on Jammable was trained by someone in the community. To make one, a person uploads or records at least 3 minutes of a single, clean voice (6 to 10 minutes sounds better) on the Train page, and a 10-minute dataset is usually ready in about 25 minutes. They can keep the voice private or share it in the public library, which is how the anime, cartoon and gaming categories keep growing.

You can do the same: train your own voice, or a voice whose owner has agreed, and use it for covers and text-to-speech like any other. Our guide on how to train your own AI voice model walks through it. Training is included on the Creator plan, and you can also upload RVC models you trained elsewhere.

How to share your character covers

Every cover gets its own page with a link you can send to anyone, and the link shows a preview card when you post it. You can also download the audio. On Creator you can turn a cover into a Marble Run video: the voices race, the one in the lead sings, and the lyrics are on screen, ready to post on TikTok and Reels.

Does it cost anything?

You can browse every character voice, listen to demos and set up a cover for free. Making covers and text-to-speech needs a plan: Basic is $1.99 for the first month, then $8.99 a month, with unlimited covers. See pricing for every plan.