Language: English

AI voice: turn your script into narration or dialogue

AI voice in shuut reads a written script aloud as narration. You paste the text, pick a voice model and a voice, and get an audio clip back.

Four voice models sit side by side in the sound tool. Each one has its own set of voices, languages and ways to steer emotion or tone. A dialogue mode lets several speakers share one clip.

Open the tool on shuut

Models for this

ModelMade byPriceLowest price
Eleven v4ElevenLabs4 credits per generation4 credits per generation
Chirp 3 HD voiceGoogle2 credits per generation2 credits per generation
MiniMax Speech 2.8 HDMiniMax5 credits per generation5 credits per generation
Qwen Audio 3.1 TTSAlibaba2 credits per generation2 credits per generation

Prices are in credits (1 credit = US$0.01) and are deducted from your shuut balance per use. You see the exact price before each generation.

How AI voice works in the shuut sound tool

The AI voice button opens the sound tool with voice already selected.

You type or paste the script that should be read.

You choose a voice model, then a voice from that model's list.

For a conversation, dialogue mode lets you assign a voice to each speaker so the whole exchange comes out as one clip.

Pressing “Generate” turns the script into narration you can play and download.

The Chirp 3 HD voices also read scripts inside the “AI Avatar” tool.

AI voice models in shuut and how they differ

Eleven v4 is ElevenLabs' speech model. It reads with emotion, offers many voices and covers a wide range of languages, and you can add tone tags to the script.

Eleven v4 Dialogue turns a written dialogue into one audio clip, with a separate voice picked for each speaker.

Chirp 3 HD is Google's text-to-speech model, focused on natural Korean narration with a small set of voices.

MiniMax Speech 2.8 HD reads Korean and English, has a broad voice list and lets you choose an emotion.

Qwen Audio 3.1 TTS has a few multilingual voices and follows tone directions you give it.

Before you write an AI voice script

Decide the language first, since it narrows the model choice: Chirp 3 HD for Korean only, MiniMax Speech 2.8 HD for Korean or English, Eleven v4 or Qwen Audio 3.1 TTS for other languages.

If the delivery matters, plan where emotion or tone should change and use Eleven v4 tone tags, the MiniMax emotion choice or Qwen tone directions.

For a dialogue, label each line by speaker so assigning voices is straightforward.

Reading in your own cloned voice is a separate Eleven v4 mode and needs the consent of the person whose voice it is.

If a real person's voice appears, check whether the place you post it needs an AI-generated label.

How to make an AI voice clip in shuut

  1. Open AI voice to go to the sound tool with voice selected.
  2. Paste your script, or write a dialogue with each line marked by speaker.
  3. Pick a voice model and a voice, or one voice per speaker in dialogue mode.
  4. Add emotion, tone tags or tone directions where the model supports them.
  5. Press “Generate” and play back the finished narration.

Frequently asked questions

Which languages can AI voice read?
Eleven v4 covers a wide range of languages, and Qwen Audio 3.1 TTS has multilingual voices. MiniMax Speech 2.8 HD reads Korean and English, and Chirp 3 HD reads Korean.
Which models have Korean voices?
Chirp 3 HD is built for natural Korean narration, and MiniMax Speech 2.8 HD reads Korean as well as English. Eleven v4 and Qwen Audio 3.1 TTS are multilingual options too.
Can one clip have more than one speaker?
Yes. In dialogue mode you pick a voice for each speaker, and Eleven v4 Dialogue turns the whole conversation into a single audio clip with several speakers.
How do I make the voice sound happy, calm or serious?
Eleven v4 reads with emotion and accepts tone tags in the script. MiniMax Speech 2.8 HD has an emotion choice, and Qwen Audio 3.1 TTS follows tone directions.
Can it read in my own voice?
Reading in a cloned voice is a separate Eleven v4 mode. It needs the consent of the person whose voice is cloned.
How is an AI voice clip priced?
Voice jobs are priced per run. Each model shows its price as a quote before you press “Generate”.

What you can make

Open the tool on shuut

Updated November 17, 2026

Before you start

You pay per result, with a quote before you press “Generate”; subscriptions are optional.

Motion shots, captions and resizing are free once you sign up.

Section 9 of the Terms covers commercial use; scope varies by model and agreement, and checking rights, ad rules and AI labels is up to you.

shuut isn't affiliated with the model makers.

Prices · Terms