You can connect your own ElevenLabs account and add any voice it has access to as a custom voice, including cloned voices. Audio made with a custom voice uses your ElevenLabs credits, not Open Meadow credits, so custom voices work on every plan, including Free.
Custom voices are separate from Open Meadow's built-in ElevenLabs voices. The built-in voices use Open Meadow credits, and you don't need an ElevenLabs account for them.
Requirements
- An ElevenLabs account.
- An ElevenLabs API key with the Text to Speech permission.
- The voice ID of each voice you want to use.
Most ElevenLabs library voices require at least the ElevenLabs Starter plan. On the ElevenLabs free plan, you can only use their default voices.
Connect your account
- In ElevenLabs, open API keys and create a restricted key with Text to Speech enabled.
- In Open Meadow, open Preferences and find Custom Voices.
- Choose Connect ElevenLabs Account, paste the key into API Key, and choose Connect.
When you connect, Open Meadow checks the key with a short speech test. If the test fails, the dialog says why, for example a missing Text to Speech permission, too few ElevenLabs credits, or IP restrictions on the key.
Open Meadow stores the key encrypted. Disconnect under Custom Voices deletes the key immediately and makes your custom voices unavailable. They are not deleted, and they come back when you connect a key again.
Add a voice
- Under Custom Voices, choose Add Voice.
- In ElevenLabs, open the voice you want and copy its voice ID.
- Fill in Voice Name and Voice ID, and choose a Model.
- Adjust the voice settings if needed.
- Enter a sentence under Test to hear the voice. Each test uses your ElevenLabs credits.
- Choose Create Voice.
Instead of entering a voice ID, you can start from one of the voices in Preset. Saved voices appear in the Custom Voices list, where you can edit or delete them.
Voice settings
| Setting | Range | Default |
|---|---|---|
| Model | Eleven v4, Eleven v4 Turbo, Eleven v3, Multilingual v1, Multilingual v2, Flash v2, Flash v2.5, Turbo v2, Turbo v2.5 | Multilingual v2 |
| Speed | 0.70 to 1.20 | 1.00 |
| Stability | 0 to 1 | 0.50 |
| Similarity Boost | 0 to 1 | 0.75 |
| Style | 0 to 1 | 0 |
| Use Speaker Boost | On or off | Off |
| Volume | -12 dB to +6 dB | 0 dB |
Volume is an Open Meadow setting. It makes the voice quieter or louder after ElevenLabs generates it, which helps when a voice is too loud or too quiet for a guided meditation. The other settings are ElevenLabs settings. Open Meadow sends them to ElevenLabs each time it generates speech with the voice.
When changed settings take effect
New settings apply to speech generated after you save them. Meditations you already played with the voice keep the audio they were made with. To hear one with the new settings, open it in the player, choose More, then Other Settings, then Resynthesize. This generates the speech again and uses your ElevenLabs credits. See Fix audio problems.
Use a custom voice
- For one meditation: open the meditation card's actions menu, choose Voice and Music, and select the voice in Playback Settings.
- For new meditations: open Preferences and select the voice under Default Player Settings.
Changing the voice of a saved meditation generates new speech audio with that voice, using your ElevenLabs credits.
In Playback Settings, custom voices appear in their own group, Your custom voices, for meditations in every language.
Use your own voice
To hear meditations in your own voice, first create a clone of it in ElevenLabs. Then add the cloned voice with its voice ID, as described above. Cloning a voice needs its owner's permission.