> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.mythos.new/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.mythos.new/_mcp/server.

# Connect ElevenLabs for voice and audio

> Add speech, transcription, sound effects, music, and voice agents to your Mythos app with ElevenLabs.

[ElevenLabs](https://elevenlabs.io/) provides tools for creating and working with audio. Connect your account to Mythos to turn text into speech, transcribe recordings, generate sound effects and music, or add a conversational voice agent.

ElevenLabs is an **App + chat** connector. You can use it while building in project chat and in your published app. Audio processing uses the account and permissions behind your saved API key.

Use it when listening or speaking is part of your app: narrated lessons, spoken summaries, accessible reading, recorded interviews, or a voice assistant. You can choose from voices available in your ElevenLabs account.

## Examples to build

| App idea              | Prompt to try                                                                          | What it adds                                              |
| --------------------- | -------------------------------------------------------------------------------------- | --------------------------------------------------------- |
| Spoken assistant      | “Build an assistant that reads its answers aloud using my selected ElevenLabs voice.”  | Play a spoken version of each answer.                     |
| Audio briefing        | “Turn the daily summary in my app into an ElevenLabs audio briefing.”                  | Let users listen to the same information shown on screen. |
| Narrated articles     | “Add a Listen button to each article using ElevenLabs.”                                | Generate narration and provide playback beside the text.  |
| Story player          | “Create a story player with different ElevenLabs voices for its characters.”           | Produce narrated dialogue for the story's characters.     |
| Recording transcripts | “Transcribe uploaded meeting recordings with ElevenLabs and show speaker labels.”      | Turn an MP3 or WAV recording into a readable transcript.  |
| Spoken lessons        | “Read each lesson and exercise aloud with ElevenLabs.”                                 | Add audio instruction to a learning flow.                 |
| Guided sessions       | “Build a stretching session with spoken instructions from ElevenLabs.”                 | Guide users through a sequence with audio cues.           |
| Accessible reading    | “Add an ElevenLabs reader for the content on this page, with play and pause controls.” | Offer an alternative way to consume the page's text.      |

For an interactive conversation, specify the existing ElevenLabs agent you want to use. For generated audio, include the intended voice, language, and whether the app should keep the audio for future playback.

## Connect ElevenLabs to Mythos

You need permission to create **App + chat** connections in your workspace. If a connection is already available to you, link that connection to your project.

### Before you start

Have an ElevenLabs account, an API key, and credits for the features you plan to use. The key must allow **user read access** for the connection check, as well as the audio features your app needs.

If your account uses a data residency environment, you also need to know its region. A key from one environment does not automatically work in another.

ElevenLabs bills audio generation, transcription, and other paid features under its own plan. The connection does not add ElevenLabs credits to your account.

### 1. Create a key in ElevenLabs

1. Sign in and open **Developers → [API keys](https://elevenlabs.io/app/developers/api-keys)**.
2. Create a key with a clear name, such as `Mythos audio`.
3. Restrict the key to the features you want the app to use. Keep user read access enabled so Mythos can check the account.
4. Set a credit limit for the key if you want to limit its usage.
5. Copy the key and store it securely.

Do not paste it into project chat or your app's public code. See the [ElevenLabs authentication guide](https://elevenlabs.io/docs/api-reference/authentication) for key permissions.

### 2. Add the connection in Mythos

1. From the Dashboard, open **Connectors** and select **ElevenLabs**. If a connection already exists, select **Add connection** to open a new form.
2. Enter a **Display name**, such as `Learning app audio`.
3. Paste your key into **API key**.
4. Choose the matching **API host**: **Default**, **Europe residency**, **India residency**, or **Singapore residency**.
5. Under **Sharing**, keep **Private** or choose **Share with others**. Use **Share with workspace** for access across the workspace.
6. Click **Connect ElevenLabs**. Mythos checks the account without generating audio.

Open your project, choose **Chat actions (+) → Project connectors**, select the saved connection, and click **Use in this project**.

Start by asking:

> Show the ElevenLabs voices available to this connection. Help me choose one for short English lessons, then add audio playback to the first lesson.

Use a short sample for the first generation. Check that the app plays the audio, that the voice and language are right, and that the output is saved when you want to reuse it.

## Limitations

* **The selected key controls access.** A key allowed to read voices may still be unable to generate speech. Enable the relevant feature in ElevenLabs if you intend to use it.
* **Voice cloning and new dubbing jobs are not set up through this connector.** Prepare these in ElevenLabs. You can then work with supported existing resources.
* **The connection does not rotate its own key.** Create a new connection after changing the key in ElevenLabs.
* **Linked projects use your account.** This connection does not give every visitor a separate ElevenLabs login or credit balance.

If the connection check fails, confirm the key belongs to the selected API host and has user read access. If audio generation fails later, check the key's feature permissions, credit limit, available credits, and the chosen voice.

## Manage the connection

Open **Connectors → ElevenLabs** and select the saved connection. **Unlink projects** removes access for selected projects; **Disconnect** stops access through that connection.

Existing saved audio is not deleted when you disconnect. Separately configured agents or calls in ElevenLabs are not automatically removed. To replace the key, disconnect the old connection, create a new one with the replacement key, and link it to the affected projects. Ask Mythos to set up their app features again.