Skip to content
verify mcp Beta VerifyMCP is currently in beta. If you notice any issues, get in touch and we’ll put it right.

com.mcparmory/elevenlabs

PYPI · MCPARMORY-ELEVENLABS · 2 COMPONENTS · SCANNED SEP 20

Convert text to speech, transcribe audio, and dub videos with AI voices

0 this week 75 Trust /100
Trust breakdown (7 categories)

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →

Supply Chain Security87
  • No malware found by supply-chain analysis.Pass
  • No known CVEs affecting this package version or its production dependencies.Pass
  • Install-script risk not yet assessed.Unverified
  • 2 of 41 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency32
Schema Quality & AI Usability70
  • AI-judged instruction clarity (excellent).Pass
  • Context-footprint check failed: tool/resource definitions use about 38535 tokens (~149/item across 258 items; 258 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
  • Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management87
  • Stability observed for 26 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage100
  • 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
  • 100% of tool parameters carry a description.Pass
Tool Safety95
  • No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
  • 32 of 40 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "isolate_audio" implies "remove" and declares no destructiveHint at all, which the MCP spec reads as destructive by default. See how to fix → Partial
  • An AI judge read all 258 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
  • Implements a current MCP spec version (2026-07-28).Pass
Install

How do I install the com.mcparmory/elevenlabs MCP server?

com.mcparmory/elevenlabs runs locally as a PyPI package, launched with uvx mcparmory-elevenlabs. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

pypi · mcparmory-elevenlabs

# add to Claude Code
claude mcp add com-mcparmory-elevenlabs -- uvx mcparmory-elevenlabs
// .cursor/mcp.json
{
  "mcpServers": {
    "com-mcparmory-elevenlabs": {
      "command": "uvx",
      "args": [
        "mcparmory-elevenlabs"
      ]
    }
  }
}
// .vscode/mcp.json
{
  "servers": {
    "com-mcparmory-elevenlabs": {
      "command": "uvx",
      "args": [
        "mcparmory-elevenlabs"
      ]
    }
  }
}
# add to Codex CLI
codex mcp add com-mcparmory-elevenlabs -- uvx mcparmory-elevenlabs
// opencode.json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "com-mcparmory-elevenlabs": {
      "type": "local",
      "command": [
        "uvx",
        "mcparmory-elevenlabs"
      ],
      "enabled": true
    }
  }
}
# add to OpenClaw
openclaw mcp add com-mcparmory-elevenlabs --command uvx --arg mcparmory-elevenlabs
# ~/.hermes/config.yaml
mcp_servers:
  com-mcparmory-elevenlabs:
    command: "uvx"
    args: ["mcparmory-elevenlabs"]
// ~/.netclaw/config/netclaw.json
{
  "McpServers": {
    "com-mcparmory-elevenlabs": {
      "Transport": "stdio",
      "Command": "uvx",
      "Arguments": [
        "mcparmory-elevenlabs"
      ]
    }
  }
}
# add to Vellum
assistant mcp add com-mcparmory-elevenlabs -t stdio -c uvx -a mcparmory-elevenlabs
// mcp.json
{
  "mcpServers": {
    "com-mcparmory-elevenlabs": {
      "command": "uvx",
      "args": [
        "mcparmory-elevenlabs"
      ]
    }
  }
}
Changelog

Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.

  • 20 Sept 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 83 to 87. That category is still filling its 30-day observation window: 25 days of observed history at the previous scan, 26 at this one. The score rises as the window fills, whether or not the server changes.

  • 18 Sept 26 +12
    • Malware scan: unverified → pass security
    • Stability: pass → 0.80 functional
  • 17 Sept 26 −14
    • Malware scan: pass → unverified security
    • Stability: 0.97 → pass security
  • 15 Sept 26 +16
    • Malware scan: unverified → pass security
  • 14 Sept 26 −15
    • Malware scan: pass → unverified security
  • 13 Sept 26 +1

    No change was recorded against any check on this day. Stability & Change Management went from 83 to 87. That category is still filling its 30-day observation window: 25 days of observed history at the previous scan, 26 at this one. The score rises as the window fills, whether or not the server changes.

  • 11 Sept 26 −3
    • Stability: pass → 0.80 functional
  • 10 Sept 26 +1
    • Stability: 0.97 → pass security
Diagnostics

Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.

Captured 20 Sept 2026 · Analysed pypi/mcparmory-elevenlabs@1.0.3

Provenance No attestation

The registry publishes no build provenance for this version, so there is nothing to verify.

Result No attestation
Ecosystem pypi

Background: How many MCP packages publish verified provenance →

Dependencies 41 packages
Packages resolved 41
Stale 1
No linked repository 1
Tree resolution Complete

Background: SBOMs and build attestations, explained →

MCP tools · 258 exposed · ~38,535 tokens

The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →

Tool Tokens
delete_conversation_file ~76

Remove a file upload from a conversation. This operation is only available if the file has not yet been used within the conversation.

NameTypeReqDescription
conversation_idstringyesThe unique identifier of the conversation containing the file to be deleted.
file_idstringyesThe unique identifier of the file upload to be removed from the conversation.

No output schema declared.

No examples provided.

delete_dubbing ~46

Permanently deletes a dubbing project and all associated data. This action cannot be undone.

NameTypeReqDescription
dubbing_idstringyesThe unique identifier of the dubbing project to delete.

No output schema declared.

No examples provided.

delete_dubbing_segment ~77

Removes a single segment from a dubbing project. This operation permanently deletes the specified segment and cannot be undone.

NameTypeReqDescription
dubbing_idstringyesThe unique identifier of the dubbing project containing the segment to be deleted.
segment_idstringyesThe unique identifier of the segment to be deleted from the dubbing project.

No output schema declared.

No examples provided.

delete_evaluation_criterion ~52

Delete a specific evaluation criterion from the speech-to-text evaluation system. This operation permanently removes the criterion and cannot be undone.

NameTypeReqDescription
criterion_idstringyesThe unique identifier of the evaluation criterion to delete.

No output schema declared.

No examples provided.

delete_history_item ~60

Delete a speech history item by its ID. This removes the item from your speech synthesis history.

NameTypeReqDescription
history_item_idstringyesThe unique identifier of the history item to delete. You can retrieve available history item IDs using the list history items endpoint.

No output schema declared.

No examples provided.

delete_human_agent ~51

Remove a human agent from the speech-to-text evaluation system. This operation permanently deletes the agent and their associated routing configuration.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the human agent to delete.

No output schema declared.

No examples provided.

delete_knowledge_base_document ~127

Permanently delete a document or folder from the knowledge base. Optionally force deletion even if the document is in use by agents, which will also remove it from dependent agents and delete all child documents in non-empty folders.

NameTypeReqDescription
documentation_idstringyesThe unique identifier of the document or folder to delete from the knowledge base.
forcebooleanForce deletion of the document or folder even if it is currently used by agents. When enabled, the document will be removed from all dependent agents, and all child documents and folders within non-e…

No output schema declared.

No examples provided.

delete_mcp_server ~50

Remove a specific MCP server configuration from the workspace. This action permanently deletes the server and its associated settings.

NameTypeReqDescription
mcp_server_idstringyesThe unique identifier of the MCP server to delete.

No output schema declared.

No examples provided.

delete_phone_number ~49

Delete a phone number from your ConvAI account by its ID. This action is permanent and cannot be undone.

NameTypeReqDescription
phone_number_idstringyesThe unique identifier of the phone number to delete.

No output schema declared.

No examples provided.

delete_project ~42

Permanently deletes a Studio project and all associated data. This action cannot be undone.

NameTypeReqDescription
project_idstringyesThe unique identifier of the Studio project to delete.

No output schema declared.

No examples provided.

delete_pronunciation_rules ~92

Remove one or more pronunciation rules from a pronunciation dictionary. Specify the dictionary ID and provide the list of rule strings to be deleted.

NameTypeReqDescription
pronunciation_dictionary_idstringyesThe unique identifier of the pronunciation dictionary from which rules will be removed.
rule_stringsarrayyesAn array of rule strings to remove from the pronunciation dictionary. Each string represents a rule to be deleted. Order is not significant.

No output schema declared.

No examples provided.

delete_rag_index ~84

Delete a RAG index associated with a knowledge base document. This removes the indexed data used for retrieval-augmented generation on that document.

NameTypeReqDescription
documentation_idstringyesThe unique identifier of the knowledge base document whose RAG index will be deleted.
rag_index_idstringyesThe unique identifier of the RAG index to delete for the specified document.

No output schema declared.

No examples provided.

delete_secret ~43

Delete a workspace secret. The secret must not be in use by any active configurations before deletion.

NameTypeReqDescription
secret_idstringyesThe unique identifier of the secret to delete.

No output schema declared.

No examples provided.

delete_tool ~88

Delete a tool from the workspace. Optionally force deletion to remove the tool from all dependent agents and branches regardless of current usage.

NameTypeReqDescription
forcebooleanForce deletion of the tool even if it is currently used by agents or branches. When enabled, the tool will be automatically removed from all dependent agents and branches.
tool_idstringyesThe unique identifier of the tool to delete.

No output schema declared.

No examples provided.

delete_transcript ~41

Permanently delete a transcript by its unique ID. This action cannot be undone.

NameTypeReqDescription
transcription_idstringyesThe unique identifier of the transcript to delete.

No output schema declared.

No examples provided.

delete_voice ~51

Permanently deletes a voice by its ID. This action cannot be undone.

NameTypeReqDescription
voice_idstringyesThe unique identifier of the voice to delete. You can retrieve available voice IDs from the list voices endpoint.

No output schema declared.

No examples provided.

delete_voice_sample ~66

Permanently removes a sample from a voice by its ID. This action cannot be undone.

NameTypeReqDescription
sample_idstringyesThe unique identifier of the sample to delete from the specified voice.
voice_idstringyesThe unique identifier of the voice containing the sample to delete.

No output schema declared.

No examples provided.

delete_whatsapp_account ~53

Permanently delete a WhatsApp account and remove it from the ConvAI platform. This action cannot be undone.

NameTypeReqDescription
phone_number_idstringyesThe unique identifier for the WhatsApp phone number account to delete.

No output schema declared.

No examples provided.

deploy_agent ~79

Create or update deployments for an agent, specifying which branches to deploy and how to distribute traffic across them.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent for which to create or update deployments.
requestsarrayyesAn ordered list of deployment configurations, each specifying a branch and its traffic allocation strategy. Order may affect deployment precedence.

No output schema declared.

No examples provided.

design_voice ~260

Generate voice design previews based on a detailed description. Returns multiple voice options with audio samples that can be used to create a custom voice.

NameTypeReqDescription
loudnessnumberVolume level adjustment for the generated voice, where -1 is quietest and 1 is loudest. A value of 0 corresponds to approximately -24 LUFS.
model_idstringAI model version to use for voice generation. Different models may produce varying quality and multilingual support.
output_formatstringAudio codec, sample rate, and bitrate for the generated voice samples. Higher bitrates and sample rates require appropriate subscription tiers.
qualitynumberQuality level for voice generation, where higher values produce better output but with less variation across previews.
should_enhancebooleanAutomatically enhance the voice description with AI-generated details to improve voice generation quality and variety. Expands simple prompts into more comprehensive descriptions.
stream_previewsbooleanWhen enabled, voice previews are streamed separately via the stream endpoint instead of being included in the response. Useful for reducing response payload size.
voice_descriptionstringyesDetailed description of the desired voice characteristics. Used to guide voice generation and should include personality, tone, and acoustic qualities.

No output schema declared.

No examples provided.

download_dubbed_audio ~90

Download the dubbed audio file in MP3 or MP4 format for a specific language. Returns the original automatic dub result; for edited dubs created in Dubbing Studio, use the render endpoint instead.

NameTypeReqDescription
dubbing_idstringyesThe unique identifier of the dubbing project containing the dubbed content.
language_codestringyesThe language code specifying which dubbed audio track to retrieve.

No output schema declared.

No examples provided.

download_pronunciation_dictionary_version ~74

Download a PLS (Pronunciation Lexicon Specification) file containing the rules for a specific version of a pronunciation dictionary.

NameTypeReqDescription
dictionary_idstringyesThe unique identifier of the pronunciation dictionary to retrieve.
version_idstringyesThe unique identifier of the specific version of the pronunciation dictionary to download.

No output schema declared.

No examples provided.

download_snapshot_archive ~75

Downloads a compressed archive containing all audio files from a specific Studio project snapshot. Returns the archive as a binary stream ready for download.

NameTypeReqDescription
project_idstringyesThe unique identifier of the Studio project containing the snapshot to archive.
project_snapshot_idstringyesThe unique identifier of the project snapshot to archive and download.

No output schema declared.

No examples provided.

download_speech_items ~102

Download one or more speech history items as audio files. Single items are returned as individual audio files, while multiple items are packaged into a .zip archive.

NameTypeReqDescription
history_item_idsarrayyesList of history item IDs to download. Retrieve available IDs and metadata from the list speech history endpoint. Order is preserved in the output archive.
output_formatstringAudio file format for transcoding. Specify the desired output format for the downloaded audio files.

No output schema declared.

No examples provided.

dub_media ~601

Dubs an audio or video file into a target language with automatic speaker detection and voice synthesis. Supports advanced options for quality control, voice customization, and manual transcript editing.

NameTypeReqDescription
csv_filestringBase64-encoded file content for upload. CSV file containing transcription and translation metadata for manual dubbing mode. Used to override automatic transcription and provide custom timing and spea…
csv_fpsnumberFrames per second value to use when parsing timecodes in the CSV file. If omitted, FPS will be automatically inferred from the timecode data.
disable_voice_cloningbooleanWhether to use similar voices from the ElevenLabs Voice Library instead of cloning the original speaker's voice. Requires 'add_voice_from_voice_library' workspace permission and consumes available cu…
drop_background_audiobooleanWhether to remove background audio from the final dub. Recommended for content like speeches or monologues where background noise should not be preserved.
dubbing_studiobooleanWhether to prepare the output for editing in the dubbing studio interface or as a dubbing resource for further processing.
end_timeintegerEnd time in seconds at which to stop dubbing the source file. Useful for processing only a segment of the content.
highest_resolutionbooleanWhether to process and output the video at the highest available resolution. May increase processing time and resource usage.
modestringProcessing mode for the dubbing job. Use 'automatic' for standard processing or 'manual' when providing a custom CSV transcript. Manual mode is experimental and not recommended for production use.
namestringHuman-readable name for the dubbing project to help organize and identify the job.
num_speakersintegerNumber of distinct speakers to use in the dubbing. Set to 0 to automatically detect the speaker count from the source audio.
source_langstringLanguage code of the source content using ISO 639-1 or ISO 639-3 format. Set to 'auto' to automatically detect the language.
source_urlstringURL pointing to the source video or audio file to be dubbed. Must be publicly accessible.
start_timeintegerStart time in seconds from which to begin dubbing the source file. Useful for processing only a segment of the content.
target_accentstringOptional accent preference to apply when selecting voices and informing translation dialect. This is an experimental feature.
target_langstringLanguage code for the target dub using ISO 639-1 or ISO 639-3 format. Determines which language the content will be dubbed into.
use_profanity_filterbooleanWhether to censor profanities in transcripts by replacing them with '[censored]'. This is a beta feature.
watermarkbooleanWhether to add a watermark overlay to the output video file.

No output schema declared.

No examples provided.

duplicate_agent ~73

Create a new agent by duplicating an existing agent. The new agent will have the same configuration as the source agent, with an optional custom name.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent to duplicate.
namestringAn optional custom name for the duplicated agent to help identify it.

No output schema declared.

No examples provided.

estimate_agent_llm_cost ~174

Estimates the expected number of LLM tokens required for an agent based on prompt length, knowledge base content, and RAG configuration. Use this to forecast token consumption and associated costs before deployment.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent for which to calculate expected LLM token usage.
number_of_pagesintegerThe total number of pages in PDF documents or URLs indexed in the agent's Knowledge Base. Used to estimate token consumption for RAG retrieval and context injection.
prompt_lengthintegerThe length of the input prompt in characters. Used to estimate token consumption for the prompt component.
rag_enabledbooleanWhether Retrieval-Augmented Generation (RAG) is enabled for the agent. When enabled, additional tokens are consumed for knowledge base retrieval and context augmentation.

No output schema declared.

No examples provided.

fetch_agent_response_test_summaries ~63

Retrieve summaries for multiple agent response tests by their IDs. Returns a mapping of test IDs to their corresponding test summary data.

NameTypeReqDescription
test_idsarrayyesList of unique test IDs to retrieve summaries for. Each ID identifies a specific agent response test.

No output schema declared.

No examples provided.

find_group ~60

Searches for user groups in the workspace by name. Returns matching group(s) or an empty result if no groups are found.

NameTypeReqDescription
namestringyesThe name of the user group to search for. The search will match against group names in the workspace.

No output schema declared.

No examples provided.

find_similar_voices ~137

Find voices from the library that are similar to a provided audio sample. Returns a ranked list of matching voices based on similarity scoring.

NameTypeReqDescription
audio_filestringBase64-encoded file content for upload. Audio sample file to match against library voices. Used as the reference for similarity comparison.
similarity_thresholdnumberSimilarity threshold for filtering results. Lower values return more similar voices. Valid range is 0 to 2.
top_kintegerMaximum number of similar voices to return. If similarity_threshold is also specified, fewer voices may be returned. Valid range is 1 to 100.

No output schema declared.

No examples provided.

generate_composition_plan ~197

Generate a detailed composition plan from a text prompt, specifying musical structure, styles, and duration for music generation.

NameTypeReqDescription
model_idstringThe AI model version to use for generating the composition plan.
music_length_msintegerTarget duration for the composition in milliseconds. If omitted, the model will automatically determine an appropriate length based on the prompt.
negative_global_stylesarrayyesArray of musical styles and directions to exclude from the entire composition. Specify in English for optimal results.
positive_global_stylesarrayyesArray of musical styles and directions that should be emphasized throughout the entire composition. Specify in English for optimal results.
promptstringyesText prompt describing the desired composition, musical style, mood, and any specific creative direction.
sectionsarrayyesArray of song sections defining the structure and progression of the composition. Order matters and determines the sequence of sections in the final output.

No output schema declared.

No examples provided.

generate_dialogue ~345

Converts a list of text and voice ID pairs into multi-voice dialogue audio. Supports up to 10 unique voices per request with configurable audio format, model, stability, and text normalization settings.

NameTypeReqDescription
apply_text_normalizationstringText normalization mode: 'auto' applies normalization based on system decision, 'on' always applies it, 'off' disables it. Normalization handles cases like spelling out numbers.
inputsarrayyesArray of dialogue segments, each containing text and a voice ID. Order is preserved in the output. Maximum of 10 unique voice IDs per request.
language_codestringISO 639-1 language code to enforce language for the model and text normalization. Returns an error if the model does not support the specified language.
model_idstringModel identifier for text-to-speech conversion. Query available models via GET /v1/models and verify can_do_text_to_speech capability.
output_formatAudio output format specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). MP3 192kbps requires Creator tier or above; PCM and WAV at 44.1kHz require Pro tier or above. μ-law format is commonl…
pronunciation_dictionary_locatorsarrayList of pronunciation dictionary locators to apply in order. Each locator contains a pronunciation_dictionary_id and version_id. Maximum of 3 locators per request.
stabilitynumberVoice stability control between 0.0 and 1.0. Lower values increase emotional range and variation; higher values produce more monotonous, consistent speech.

No output schema declared.

No examples provided.

generate_dialogue_stream ~371

Converts a list of text and voice ID pairs into multi-voice dialogue speech and streams the audio. Useful for creating conversations, interviews, or multi-speaker content with different voices.

NameTypeReqDescription
apply_text_normalizationstringText normalization mode: 'auto' applies normalization automatically based on content (e.g., spelling out numbers), 'on' always applies normalization, 'off' disables it entirely.
inputsarrayyesArray of dialogue turns, each containing text to speak and the voice ID to use. Order matters—items are processed sequentially to create the dialogue flow. Maximum of 10 unique voice IDs per request.
language_codestringISO 639-1 language code to enforce language processing and text normalization. If the selected model doesn't support the specified language, an error will be returned.
model_idstringModel identifier for text-to-speech processing. The model must support text-to-speech capability. Query available models via GET /v1/models and check the can_do_text_to_speech property.
output_formatstringAudio output format specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). Some formats require higher subscription tiers: MP3 192kbps requires Creator tier or above, PCM 44.1kHz requires Pro…
pronunciation_dictionary_locatorsarrayList of pronunciation dictionary locators to apply in order. Each locator contains a pronunciation_dictionary_id and version_id. Maximum of 3 locators per request.
stabilitynumberVoice stability control between 0.0 and 1.0. Lower values increase emotional range and variability; higher values produce more consistent, monotonous speech.

No output schema declared.

No examples provided.

generate_dialogue_stream_with_timestamps ~327

Converts text and voice ID pairs into streamed dialogue audio with precise timestamps. Returns a continuous stream of JSON objects containing base64-encoded audio chunks and their corresponding timing information.

NameTypeReqDescription
apply_text_normalizationstringControls text normalization behavior. 'auto' lets the system decide, 'on' always normalizes (e.g., converts numbers to words), 'off' disables normalization.
inputsarrayyesArray of dialogue turn objects, each pairing text content with a voice ID. Processed in order to create sequential dialogue. Maximum of 10 unique voice IDs per request.
language_codestringISO 639-1 language code to enforce language processing and text normalization. The selected model must support the specified language.
model_idstringThe TTS model to use for synthesis. Query available models via GET /v1/models and verify can_do_text_to_speech capability.
output_formatstringAudio codec, sample rate, and bitrate configuration. Format is codec_sample_rate_bitrate (e.g., mp3_44100_128). Higher bitrates and PCM formats require elevated subscription tiers.
pronunciation_dictionary_locatorsarrayOrdered list of pronunciation dictionary references to apply custom pronunciations. Applied sequentially in the order provided. Maximum of 3 locators per request.
stabilitynumberControls voice consistency and emotional variation. Lower values (closer to 0) produce greater emotional range and variability. Higher values (closer to 1) produce more consistent, monotonous deliver…

No output schema declared.

No examples provided.

generate_dialogue_with_timestamps ~337

Generate dialogue from text with precise character-level timing information for audio-text synchronization. Each dialogue turn is converted to speech using specified voice IDs and returned with exact timestamp markers.

NameTypeReqDescription
apply_text_normalizationstringText normalization mode: 'auto' applies normalization based on system decision, 'on' always applies it, 'off' disables it. Normalization handles cases like spelling out numbers.
inputsarrayyesList of dialogue turns, each containing text to be spoken and the voice ID to use for that turn. Maximum of 10 unique voice IDs per request. Turns are processed in order.
language_codestringISO 639-1 language code to enforce language for the model and text normalization. If the model does not support the specified language, an error will be returned.
model_idstringThe text-to-speech model to use for generation. Must support text-to-speech capability. Query available models via GET /v1/models to verify can_do_text_to_speech property.
output_formatAudio codec, sample rate, and bitrate format. Format is specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). Higher bitrates and certain formats require higher subscription tiers.
pronunciation_dictionary_locatorsarrayCustom pronunciation dictionary rules to apply to the text in order. Each locator references a specific dictionary version. Maximum of 3 locators per request.
stabilitynumberVoice stability control affecting emotional range and consistency. Lower values produce broader emotional variation; higher values result in more monotonous, emotionally limited speech.

No output schema declared.

No examples provided.

generate_podcast ~406

Generate a podcast by converting source content into audio using AI-powered text-to-speech. Supports both conversational (two-voice dialogue) and bulletin (monologue) formats with customizable quality, duration, language, and styling options.

NameTypeReqDescription
apply_text_normalizationstringControls text normalization behavior. 'auto' lets the system decide, 'on' always normalizes, 'apply_english' normalizes assuming English text, and 'off' disables normalization.
callback_urlstringWebhook URL for conversion status notifications. The service will POST status updates when the project and chapters complete processing, including success/error details.
duration_scalestringTarget podcast length. Controls the amount of content included in the generated podcast.
highlightsarrayKey themes or highlights summarizing the podcast content. Each highlight should be a brief phrase between 10-70 characters.
instructions_promptstringCustom instructions to guide the podcast generation style, tone, and content treatment. Use this to enforce accuracy, adjust formality, or specify audience appropriateness.
introstringOptional opening text to prepend to the podcast. Useful for branding or context-setting.
languagestringTwo-letter ISO 639-1 language code for the podcast content and voice generation.
modeyesThe podcast format type. 'conversation' generates dialogue between two voices (host and guest), while 'bulletin' generates a single-voice monologue.
model_idstringyesThe voice model to use for audio generation. Query GET /v1/models to see all available models.
outrostringOptional closing text to append to the podcast. Useful for calls-to-action or sign-offs.
quality_presetstringAudio output quality level. Higher quality settings provide better audio fidelity with improved processing.
sourceyesThe content source for podcast generation. Can be a URL, text content, or other supported source formats.

No output schema declared.

No examples provided.

generate_sound ~215

Generate realistic sound effects from text descriptions using advanced AI models. Perfect for video production, voice-overs, and game audio.

NameTypeReqDescription
duration_secondsnumberTarget duration of the generated sound in seconds. If not specified, the optimal duration will be automatically determined from the text description.
loopbooleanEnable seamless looping for the generated sound effect. Only supported with the eleven_text_to_sound_v2 model.
model_idstringThe AI model to use for sound generation. Determines the quality and capabilities of the generated audio.
output_formatstringAudio codec, sample rate, and bitrate for the generated sound. Higher bitrates and sample rates require appropriate subscription tiers.
prompt_influencenumberControls how strictly the generation adheres to the text description. Higher values produce more consistent results but less variation; lower values allow more creative freedom.
textstringyesDetailed text description of the sound effect to generate. Be descriptive about the audio characteristics, environment, and desired qualities.

No output schema declared.

No examples provided.

generate_speech ~614

Converts text into natural-sounding speech using a selected voice and returns audio in your preferred format. Supports voice customization through stability, similarity, style, and speed controls, with optional pronunciation dictionaries and continuity features for multi-part audio generation.

NameTypeReqDescription
apply_language_text_normalizationbooleanEnables language-specific text normalization for improved pronunciation in supported languages. Currently only supported for Japanese. Warning: may significantly increase request latency.
apply_text_normalizationstringControls text normalization behavior. 'auto' lets the system decide, 'on' always applies normalization (e.g., spelling out numbers), and 'off' skips normalization entirely.
language_codestringISO 639-1 language code to enforce a specific language for the model and text normalization. The model must support the specified language or an error will be returned.
model_idstringThe AI model to use for speech generation. The model must support text-to-speech capability. Defaults to the multilingual v2 model.
next_request_idsarrayRequest IDs of audio samples that follow this generation. Useful for maintaining natural flow when regenerating a sample within a sequence. Maximum of 3 request IDs. Works best with the same model ac…
output_formatstringThe audio format and quality for the generated speech, specified as codec_sample_rate_bitrate. Higher bitrates and sample rates require higher subscription tiers. Defaults to MP3 at 44.1kHz with 128k…
previous_request_idsarrayRequest IDs of previously generated audio samples to maintain continuity when splitting large tasks. Improves speech flow when combining multiple generations. Maximum of 3 request IDs. Works best wit…
pronunciation_dictionary_locatorsarrayA list of pronunciation dictionary locators to apply to the text in order. Each locator contains a pronunciation_dictionary_id and version_id. Maximum of 3 locators per request.
similarity_boostnumberControls how closely the generated speech adheres to the original voice characteristics. Higher values maintain stronger voice similarity, while lower values allow more variation.
speednumberAdjusts speech playback speed. A value of 1.0 is normal speed; values below 1.0 slow down speech, while values above 1.0 speed it up.
stabilitynumberControls voice consistency and emotional range. Lower values (closer to 0) produce more varied emotional expression, while higher values (closer to 1) create more consistent but potentially monotonou…
stylenumberAmplifies the stylistic characteristics of the voice. A value of 0 applies no style exaggeration. Higher values increase style intensity but may increase latency and computational usage.
textstringyesThe text content to be converted into speech.
voice_idstringyesThe unique identifier of the voice to use for speech generation. Available voices can be retrieved from the voices endpoint.

No output schema declared.

No examples provided.

generate_speech_stream ~520

Converts text into streaming audio using a specified voice. Returns audio as a continuous stream in your chosen format, ideal for real-time playback or large content.

NameTypeReqDescription
apply_language_text_normalizationbooleanEnable language-specific text normalization for proper pronunciation. Currently supported for Japanese only. Warning: significantly increases request latency.
apply_text_normalizationstringText normalization mode: 'auto' applies normalization automatically, 'on' always applies it, 'off' disables it. Normalization handles number spelling and similar conversions.
language_codestringISO 639-1 language code to enforce language processing and text normalization. The model must support the specified language.
model_idstringThe AI model to use for speech generation. Must support text-to-speech capability. Query available models via the models endpoint.
next_request_idsarrayRequest IDs from samples that follow this generation. Maintains natural flow when regenerating a sample within a sequence. Maximum 3 IDs, best results with consistent model.
output_formatstringAudio output format specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). Higher bitrates and PCM formats require higher subscription tiers.
previous_request_idsarrayRequest IDs from previously generated samples to maintain speech continuity. Improves flow when splitting large tasks across multiple requests. Maximum 3 IDs, best results with consistent model.
pronunciation_dictionary_locatorsarrayPronunciation dictionary locators to apply custom pronunciation rules. Specified as objects with pronunciation_dictionary_id and version_id. Applied in order, maximum 3 per request.
similarity_boostnumberVoice similarity adherence between 0.0 and 1.0. Controls how closely the generated speech matches the original voice characteristics.
speednumberSpeech speed multiplier. Use 1.0 for normal speed, values below 1.0 to slow down, and values above 1.0 to speed up.
stabilitynumberVoice stability control between 0.0 and 1.0. Lower values increase emotional range and variation; higher values produce more consistent, monotonous speech.
stylenumberStyle exaggeration intensity. Amplifies the speaker's original style characteristics. Non-zero values increase computational cost and latency.
textstringyesThe text content to convert into speech.
voice_idstringyesThe voice to use for speech generation. Available voices can be retrieved from the voices endpoint.

No output schema declared.

No examples provided.

generate_speech_stream_with_timestamps ~536

Converts text to speech audio and returns a stream of JSON objects containing base64-encoded audio chunks with character-level timing information, enabling precise synchronization of audio with text.

NameTypeReqDescription
apply_language_text_normalizationbooleanEnable language-specific text normalization for improved pronunciation in supported languages. Currently only supports Japanese. Warning: may significantly increase request latency.
apply_text_normalizationstringText normalization mode: 'auto' applies normalization when appropriate, 'on' always applies it, 'off' disables it. Normalization handles conversions like spelling out numbers.
language_codestringISO 639-1 language code to enforce language-specific processing and text normalization. The model must support the specified language.
model_idstringThe model identifier for speech synthesis. The model must support text-to-speech conversion. Available models can be queried from the models endpoint.
next_request_idsarrayRequest IDs from subsequent speech samples to maintain continuity. Accepts up to 3 IDs applied in order. Useful when regenerating a sample while preserving natural flow with following content.
output_formatstringAudio output format specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). Higher bitrates and PCM formats require higher subscription tiers.
previous_request_idsarrayRequest IDs from previously generated speech samples to maintain continuity. Accepts up to 3 IDs applied in order. Improves flow when splitting large tasks across multiple requests.
pronunciation_dictionary_locatorsarrayPronunciation dictionary locators to apply custom pronunciation rules. Accepts up to 3 locators applied in order, each containing a pronunciation_dictionary_id and version_id.
similarity_boostnumberVoice similarity control ranging from 0 to 1, determining how closely the synthesis adheres to the original voice characteristics.
speednumberSpeech speed multiplier where 1.0 is normal speed, values below 1.0 slow down speech, and values above 1.0 accelerate it.
stabilitynumberVoice stability control ranging from 0 (high emotional range) to 1 (monotonous). Lower values produce more expressive speech with greater variation.
stylenumberStyle exaggeration level (0 to 1+) that amplifies the speaker's original style. Non-zero values increase computational cost and latency.
textstringyesThe text content to convert into speech audio.
voice_idstringyesThe voice identifier to use for speech synthesis. Available voices can be retrieved from the voices endpoint.

No output schema declared.

No examples provided.

generate_speech_with_timestamps ~533

Convert text to speech audio with precise character-level timing information for synchronizing audio playback with text. Returns audio file and timestamp data for each character.

NameTypeReqDescription
apply_language_text_normalizationbooleanEnable language-specific text normalization for proper pronunciation. Currently supported for Japanese only. Warning: may significantly increase request latency.
apply_text_normalizationstringText normalization mode: 'auto' applies normalization automatically, 'on' always applies it, 'off' disables it. Normalization handles conversions like spelling out numbers.
language_codestringISO 639-1 language code to enforce language processing and text normalization. The model must support the specified language.
model_idstringThe AI model identifier to use for text-to-speech conversion. The model must support text-to-speech capability.
next_request_idsarrayRequest IDs of subsequent speech samples to maintain continuity. Useful for regenerating a sample while preserving natural flow with following audio. Maximum of 3 request IDs. Results are best when u…
output_formatstringAudio output format specified as codec_sample_rate_bitrate (e.g., mp3_44100_128). Higher bitrates and certain formats require higher subscription tiers.
previous_request_idsarrayRequest IDs of previously generated speech samples to maintain continuity. Used when splitting large tasks across multiple requests. Maximum of 3 request IDs. Results are best when using the same mod…
pronunciation_dictionary_locatorsarrayList of pronunciation dictionary locators to apply to the text in order. Each locator contains a pronunciation_dictionary_id and version_id. Maximum of 3 locators per request.
similarity_boostnumberVoice similarity control ranging from 0 to 1. Higher values make the AI adhere more closely to the original voice characteristics.
speednumberSpeech speed multiplier where 1.0 is normal speed, values below 1.0 slow down speech, and values above 1.0 speed it up.
stabilitynumberVoice stability control ranging from 0 (high emotional range) to 1 (monotonous). Lower values produce more expressive speech with greater variation.
stylenumberStyle exaggeration level for the voice. Non-zero values amplify the speaker's style but increase computational resources and latency.
textstringyesThe text content to convert into speech audio.
voice_idstringyesThe voice identifier to use for speech generation. Available voices can be retrieved from the voices endpoint.

No output schema declared.

No examples provided.

generate_voice_previews ~210

Generate multiple voice preview samples based on a text description to help you select a custom voice. Each preview includes a unique voice ID and audio sample that can be used to create the final voice.

NameTypeReqDescription
loudnessnumberVolume level of the generated voice samples, ranging from quietest to loudest. A value of 0 corresponds to approximately -24 LUFS.
output_formatstringAudio codec, sample rate, and bitrate for the generated preview samples. Higher bitrates and sample rates provide better quality but require higher subscription tiers.
qualitynumberVoice quality level that balances output fidelity with variety. Higher values produce more consistent, polished voices with less variation across previews.
should_enhancebooleanAutomatically expand and refine the voice description using AI to add detail and improve generation quality. Useful for simple or brief descriptions.
voice_descriptionstringyesDetailed description of the desired voice characteristics. The system uses this to generate voice previews matching your specifications.

No output schema declared.

No examples provided.

get_agent ~97

Retrieve the configuration and settings for a specific agent. Optionally specify a particular version or branch to retrieve.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent to retrieve.
branch_idstringThe specific branch of the agent to retrieve. If not provided, the default branch is used.
version_idstringThe specific version of the agent to retrieve. If not provided, the latest version is used.

No output schema declared.

No examples provided.

get_agent_analytics ~109

Retrieve analytics data for a specific human agent in the speech-to-text evaluation system. Optionally filter results by creation date range.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the human agent for which to retrieve analytics.
created_afterstringFilter to include only analytics created on or after this date. Specify in ISO 8601 format.
created_beforestringFilter to include only analytics created on or before this date. Specify in ISO 8601 format.

No output schema declared.

No examples provided.

get_agent_branch ~58

Retrieve detailed information about a specific agent branch, including its configuration and settings.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent that contains the branch.
branch_idstringyesThe unique identifier of the branch to retrieve.

No output schema declared.

No examples provided.

get_agent_conversation_signed_link ~110

Generate a signed URL to initiate a conversation with an authorized agent. The signed URL provides secure access to start a new conversation session.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent with which to start the conversation.
branch_idstringThe specific branch variant of the agent to use for the conversation.
include_conversation_idbooleanWhether to include a unique conversation ID in the response. When enabled, the signed URL can only be used once.

No output schema declared.

No examples provided.

get_agent_share_link ~50

Retrieve the shareable link for an agent that can be used to share the agent with others.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent for which to retrieve the share link.

No output schema declared.

No examples provided.

get_agent_test ~48

Retrieves a specific agent response test by its ID. Use this to fetch details about a previously created test.

NameTypeReqDescription
test_idstringyesThe unique identifier of the agent response test to retrieve.

No output schema declared.

No examples provided.

get_agent_widget_config ~81

Retrieve the widget configuration for a specific agent, including settings needed to embed or display the agent's conversational interface.

NameTypeReqDescription
agent_idstringyesThe unique identifier of the agent whose widget configuration you want to retrieve.
conversation_signaturestringAn optional expiring token that enables WebSocket conversation initiation. Generate tokens using the conversation signed URL endpoint.

No output schema declared.

No examples provided.

get_audio_native_settings ~56

Retrieve player settings and configuration for an Audio Native project. Use this to access the current settings applied to a specific project.

NameTypeReqDescription
project_idstringyesThe unique identifier of the Studio project for which to retrieve Audio Native settings.

No output schema declared.

No examples provided.

Common questions

What is the com.mcparmory/elevenlabs MCP server?

com.mcparmory/elevenlabs is an MCP server listed in the public MCP registry as com.mcparmory/elevenlabs. Convert text to speech, transcribe audio, and dub videos with AI voices. This page covers its PyPI package (mcparmory-elevenlabs).

Is the com.mcparmory/elevenlabs MCP server safe to use?

com.mcparmory/elevenlabs scores 75 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 20 September 2026. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

What tools does the com.mcparmory/elevenlabs MCP server expose?

com.mcparmory/elevenlabs exposes 258 tools: list_speech_history, get_speech_history_item, delete_history_item, get_speech_history_audio, download_speech_items, and 253 more. Their descriptions and schemas cost roughly 38,535 tokens of context every time the server is loaded.

Is the com.mcparmory/elevenlabs MCP server still maintained?

com.mcparmory/elevenlabs is still listed as active in the MCP registry. We last reached this channel on 20 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.