Voices
Text to speech and the voice changer use a voice. The Voice menu of a Generate audio node lists the default ElevenLabs voices plus your organization’s own voices. Other organizations’ voices are never listed.
Custom voices (added from the library, designed or cloned) belong to your organization and use a voice slot.
| Plan | Custom voice slots |
|---|---|
| No plan, Go | 3 |
| Starter | 5 |
| Pro | 10 |
Open the voice studio
Section titled “Open the voice studio”In a Generate audio node set to Text to speech or Voice changer, press More voices. It shows the slots left. Four tabs:
Library
Section titled “Library”- Type a search and/or filter by language, gender, accent and use case.
- Press the play button to hear a voice.
- Press Use. The voice is added to your organization and selected in the node.
Design
Section titled “Design”- Describe the voice in at least 20 characters. Example: A warm, calm woman in her forties with a light Italian accent.
- Press Generate previews and listen.
- Type a name and press Save on the preview you like. Cost: $0.08 (about 10 credits).
Instant clone from a recording of your own voice, or of a speaker who explicitly agreed. The tab is hidden in uncensored projects.
- Name the voice.
- Choose This is my own voice or The speaker gave me explicit consent; for the second, type the speaker’s full name.
- Record 60 to 180 seconds in the recorder (a script to read is provided). Background noise is removed.
- Tick the confirmation that you have the right to clone this voice, then press Create voice. Cost: $0.10 (about 12 credits).
The recording is deleted once the voice exists.
My voices
Section titled “My voices”Lists your custom voices (cloned or designed). Use selects one, Delete removes it after a confirmation; nodes using it stop speaking with it.
When all slots are used, Design and Clone are disabled until you delete a voice.
Agents and MCP
Section titled “Agents and MCP”| Tool | What it does |
|---|---|
list_voices | Your organization’s custom voices (free) |
voice_library | Search the library: search, language, gender, accent, useCase, page (free) |
use_library_voice | Add a library voice: ownerId, voiceId, name |
design_voice | Design previews from a description (spends credits) |
save_voice | Keep a preview: generatedVoiceId, name, description |
clone_voice | name, consentBasis, speaker?, attested: true, seconds 60 to 180, sampleBase64 up to 3 MB (spends credits; only with the user’s consent) |
delete_voice | id |
Over REST, GET /api/v1/org/audio/voices?org=<org> answers { voices, custom }.