> For the complete documentation index, see [llms.txt](https://docs.acestudio.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.acestudio.ai/ai-tools/speech-generator.md).

# Speech Generator

Turn a script into speech or multi-speaker dialogue with pre-made, described, or reference-based voices in ACE Studio.

## What Is Speech Generator

Speech Generator turns text into speech. Add one or more speakers, write what each of them says, and the generated speech is placed on the Canvas as an audio clip at the playhead.

## How to Use Speech Generator

<figure><img src="https://3596338706-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fsn71WYzecYYQw3KkF40b%2Fuploads%2Fgit-blob-20671e5e7b303b209bd2aec24b44fcaa6446e7c6%2Fspeech-generator.png?alt=media" alt="The Speech Generator panel."><figcaption></figcaption></figure>

1. Place the playhead where the speech should start.
2. Click the <i class="fa-comment-dots">:comment-dots:</i> **Speech Generator** icon in the main toolbar.
3. Choose a model.
4. Add speakers and write the script.
5. Click <kbd>Generate</kbd>. The new audio clip is inserted at the playhead.

Speech Generator does not use a time selection. The length of the result follows the script.

### Choose a Model

Pick a model from the menu next to <kbd>Generate</kbd>.

<table><thead><tr><th width="200"></th><th>SeedAudio 1.0</th><th>ElevenLabs V3</th></tr></thead><tbody><tr><td>Speakers</td><td>Up to 3, optional</td><td>1 – 4, at least one required</td></tr><tr><td>Speaker sources</td><td>Pre-made voice, Describe a voice, Reference audio</td><td>Pre-made voice</td></tr><tr><td>Script</td><td>Free text with @speaker mentions</td><td>One block per speaker turn</td></tr><tr><td>Settings</td><td>Speed, Pitch, Loudness</td><td>Stability</td></tr><tr><td>Script length</td><td>3000 characters</td><td>2000 characters</td></tr></tbody></table>

## SeedAudio 1.0

### Add Speakers

Click **+ Speaker** and choose a source:

* **Select a voice** – pick a pre-made voice from the Voice Library.
* **Describe a voice** – enter a speaker name (up to 50 characters) and a description of the voice (up to 300 characters), such as "A grounded, non-regional voice that commands attention with calm authority." Hover over the card and click the edit button to change it later.
* **Upload reference audio** – choose **Select from canvas** to drag out a range of up to 30 seconds on the Canvas, or **Upload file** to pick an audio file of up to 30 seconds and 10 MB. The voice in the reference is used for this speaker.

You can add up to three speakers. Hover over a speaker card and click the delete button to remove it.

### Write the Script

When you add a speaker, "@Speaker says" is inserted at the cursor. Write the line after it. Type `@` anywhere in the script to mention one of your speakers, and switch speakers as often as you like.

The script is plain language, so you can also describe how a line is delivered, for example "@Jamie smiled and said, “Happy music making!”".

Mentions are ordinary text: you can edit or delete them. Removing a speaker turns its mentions into plain text. Renaming a described speaker updates its mentions.

You can also generate without any speaker. The script is then read aloud word for word in a default voice.

### Settings

Open the settings menu at the bottom left of the panel. Double-click a slider to reset it to its default.

* **Speed** – from x0.5 to x2.
* **Pitch** – shift the pitch by up to ±12 semitones.
* **Loudness** – from x0.5 to x2.

## ElevenLabs V3

<figure><img src="https://3596338706-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fsn71WYzecYYQw3KkF40b%2Fuploads%2Fgit-blob-12ab3b3fc7096f97cf747a9ee62c8089011336e2%2Fspeech-generator-elevenlabs.png?alt=media" alt="Multi-speaker dialogue with ElevenLabs V3."><figcaption></figcaption></figure>

### Add Speakers

Click **+ Speaker** to pick a voice from the Voice Library. You can add up to four speakers, and need at least one to generate.

### Write the Dialogue

The script is organized in turns. Each turn has a speaker capsule and its own text. Click **Add speaker** below the dialogue to add a turn and choose who speaks it. To change the speaker of a turn, click into its text and then click a speaker card at the top.

Audio tags in square brackets, such as `[laugh]`, `[whispers]`, or `[slowly]`, shape the delivery of the line. Closed tags are highlighted in the script.

### Settings

* **Stability** – higher values give a steadier, more consistent delivery; lower values allow more expressive variation.

## Voice Library

<figure><img src="https://3596338706-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fsn71WYzecYYQw3KkF40b%2Fuploads%2Fgit-blob-058be877e6d95dfe96e374e522e1526977b12689%2Fspeech-generator-voice-library.png?alt=media" alt="The Voice Library in Speech Generator."><figcaption></figcaption></figure>

The Voice Library lists the pre-made voices of the model you are using.

* **Discover** – browse all voices. Use the search box and the filters to narrow the list.
* **Recent** – the last 10 voices you used with this model.

Click the headphones icon to preview a voice. Click a voice to add it as a speaker and close the library.

## Generate

Click <kbd>Generate</kbd>. The result is inserted at the playhead as a regular audio clip, which you can edit like any other.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.acestudio.ai/ai-tools/speech-generator.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
