ElevenLabs MCP server for AI agents
Turn text into speech, sound effects and music, transcribe audio, clean up recordings, and manage voices and generated audio in an ElevenLabs account. Audio comes back as base64.
Runs with your users' own ElevenLabs accounts. Credentials stay on prompt-tm.com.
22 tools · 0 triggers
- Auth
- API key
- Tools
- 22
- Triggers
- 0
- Category
- Media, AI
Tools and triggers
Get the user
ELEVENLABS_GET_USER_INFO
Returns the account owner: user and workspace IDs, first name, a preview of the API key, seat type and the subscription.read-only
Get the subscription
ELEVENLABS_GET_USER_SUBSCRIPTION_INFO
Returns the plan: tier, characters used and the limit, voice slots used and the limit, when the character count resets (`next_character_count_reset_unix`) and billing status.read-only
List voices
ELEVENLABS_GET_VOICES
Lists the voices in the account with their IDs. Stops working once the workspace has more than 500 voices; use Search the voice library for shared voices.read-only
Get a voice
ELEVENLABS_GET_VOICE_BY_ID
Returns metadata about a specific voice.read-only
Get voice settings
ELEVENLABS_GET_VOICE_SETTINGS
Returns the stored settings of a voice, such as stability and similarity boost.read-only
List models
ELEVENLABS_GET_MODELS
Lists the available models. Use a model's ID as `model_id` in the speech tools.read-only
Convert text to speech
ELEVENLABS_TEXT_TO_SPEECH_FULL
Reads text aloud in a chosen voice and returns the audio as base64 (`data_base64`, with `content_type`). The default format is MP3; set `output_format` for others. Audio over 2 MB comes back without data.
Generate a sound effect
ELEVENLABS_SOUND_GENERATION
Makes a sound effect from a text description, 0.5 to 30 seconds long, and returns the audio as base64.
Convert a dialogue to speech
ELEVENLABS_TEXT_TO_DIALOGUE
Turns a list of text and voice ID pairs into one spoken conversation and returns the audio as base64.
List generated audio
ELEVENLABS_GET_SPEECH_HISTORY
Lists past text-to-speech, voice conversion, Studio and dubbing generations, newest first by default (`sort_direction` `desc`). Music and sound effects are not included.read-only
Get a generated item
ELEVENLABS_GET_SPEECH_HISTORY_ITEM_BY_ID
Returns one generation's history record. Get the audio with Get generated audio.read-only
Get generated audio
ELEVENLABS_GET_AUDIO_FULL_FROM_SPEECH_HISTORY_ITEM
Returns the audio of a past generation as base64.read-only
Delete a generated item
ELEVENLABS_DELETE_SPEECH_HISTORY_ITEM
Deletes one generation and its audio from the history.destructive
Transcribe audio
ELEVENLABS_SPEECH_TO_TEXT
Transcribes an audio or video file, sent as base64 in `file` or fetched from `source_url`, with optional speaker labels and timestamps.
Get a transcript
ELEVENLABS_GET_TRANSCRIPT_BY_ID
Retrieve a previously generated transcript by its ID.read-only
Compose music
ELEVENLABS_GENERATE
Composes a song from a text prompt or a composition plan and returns the audio as base64.
List pronunciation dictionaries
ELEVENLABS_GET_PRONUNCIATION_DICTIONARIES_METADATA
Lists the pronunciation dictionaries you can use, with their IDs and metadata.read-only
Search the voice library
ELEVENLABS_GET_LIBRARY_VOICES
Searches the shared voice library by language, accent, gender, age, use case or text.read-only
List Studio projects
ELEVENLABS_GET_PROJECTS
Lists the Studio projects with their IDs and names.read-only
Get a Studio project
ELEVENLABS_GET_PROJECT_BY_ID
Returns one Studio project with its details.read-only
Delete a voice
ELEVENLABS_DELETE_VOICE
Deletes a voice from the account. Anything still using its ID stops working.destructive
Remove background noise
ELEVENLABS_AUDIO_ISOLATION
Strips background noise from an audio file (sent as base64 in `audio`) and returns the cleaned audio as base64. Files over 3 MB are refused.
Try ElevenLabs now
Sign in, connect your account and run any of these in the playground, or ask your agent through the router.
- “Get the user”ELEVENLABS_GET_USER_INFO
- “Get the subscription”ELEVENLABS_GET_USER_SUBSCRIPTION_INFO
- “List voices”ELEVENLABS_GET_VOICES
Connect via MCP
Create an MCP server for ElevenLabs in the dashboard, then add its URL to your client. This is the Claude Code command; the client guide has the others.
claude mcp add --transport http elevenlabs \
"https://www.prompt-tm.com/mcp/<server id>?user_id=<your user>" \
--header "x-api-key: ptm_…"Connect via REST
Run a tool with one request. The response is { data, error, successful, log_id } for every tool.
curl -X POST https://www.prompt-tm.com/api/v3/tools/execute/ELEVENLABS_GET_USER_INFO \
-H "x-api-key: $PROMPT_TM_KEY" -H "content-type: application/json" \
-d '{ "user_id": "alice", "arguments": {} }'