Canonical page: https://mixerai.org/en/knowledge/audio/elevenlabs-tts/

[Mixer AI](https://mixerai.org/) › [Knowledge Base](https://mixerai.org/en/knowledge/) › [Audio](https://mixerai.org/en/knowledge/audio/) › ElevenLabs

# ElevenLabs

Voice a script with one speaker or build a dialogue with a separate voice for each line.

[Launch ElevenLabs in Mixer AI](https://mixerai.org/?dl=kb_elevenlabs-tts#/canvas)

## When to choose it

ElevenLabs in Canvas turns prepared text into speech. Choose one voice for narration or Dialogue mode for a scene. Approve the wording and name pronunciations first: changing the script after synthesis means generating that passage again.

- Ready-made voices and delivery controls for single-speaker narration.

- Dialogue v3 assigns lines to different voices.

## Modes and settings

- Multilingual v2 and Turbo 2.5 use one voice, up to 5000 characters. Turbo provides explicit language selection.

- Dialogue v3 supports up to 64 lines and 5000 characters in total, with a voice chosen for each line.

## Price in coins

| Version | Input | Coins / 1,000 characters
| Turbo v2.5 | Single voice | 3.87
| Multilingual v2 | Single voice | 7.74
| Dialogue v3 | Multiple voices | 9.03

Rates are per 1,000 characters, including spaces and punctuation. Cost = rate × character count / 1,000. Short text is not rounded up to a 1,000-character block; only the final coin total is rounded up.

≈ marks a fractional rate rounded to two decimal places for display; billing uses full precision.

## What you can make

- Narration for instructions or a promotional video.

- A short multi-speaker scene or dialogue.

## How to direct the result

- Write numbers, abbreviations and unfamiliar names as they should sound.

- Use punctuation to guide pauses. Test a short excerpt before voicing the full script.

### Example: calm instructional narration
Copy

```text
Place the cup on a level surface. Pour hot water up to the mark and wait one minute. Then stir slowly. Your drink is ready to enjoy.
```

## What to keep in mind

- The dialogue limit is shared, not 5000 characters for every speaker.

- Pronunciation errors and long sentences may require rewriting and regeneration.

## FAQ

How do I select several voices?

Use Dialogue v3 and assign a voice to each line.

Is the five-thousand-character limit per line?

No. In Dialogue it covers all lines together; in single-speaker modes it covers the one text.

Is pricing based on audio minutes?

No, it uses the actual text character count. The table gives a rate per 1,000 characters for each version; that is a comparison unit, not a minimum purchase.

Can I put 'speak cheerfully' before the script?

In a single-speaker mode, instructions inside the text may be read aloud. Use the available voice controls and keep the spoken text ready for reading.

Why is a name pronounced incorrectly?

Try a clearer phonetic spelling and test it in a short passage.

Can I upload my voice for cloning?

This page uses the ready-made voices in its menu. It does not offer voice cloning through an ordinary audio upload.

What changes the price?

It depends on the speech version and character count, including spaces and punctuation. The table gives a rate per 1,000 characters; you pay for the actual count, not whole thousand-character blocks.

Do I need a separate model subscription?

No. In Mixer AI, you top up one coin balance and use it for the tasks you need. You do not need a separate subscription to this model.

[Launch ElevenLabs in Mixer AI](https://mixerai.org/?dl=kb_elevenlabs-tts#/canvas)