LMNT focuses on voice and audio manipulation, possibly leveraging AI to generate or transform sound for various creative and technical use cases
LMNT focuses on voice and audio manipulation, possibly leveraging AI to generate or transform sound for various creative and technical use cases On Nagent, LMNT is exposed as a fully-configurable artificial intelligence integration that any agent can call — 8 actions, and API key authentication. No code is required to wire LMNT into your workflow — connect it once via the External Integrations panel and reuse it across every agent you build.
Agent builders use LMNT to automate the kinds of tasks artificial intelligence teams previously handled manually. Concrete examples — each one is a single agent step in Nagent — include:
Every action and trigger is paired with a structured input/output schema (visible in the sections below), so when you wire LMNT into Helix — our agentic agent builder — the editor knows exactly what each step expects and produces. Configure once, deploy anywhere across your Nagent agents.
Every operation an agent can call against LMNT, with input parameters and output schema. Drop these into any step of an agent built in Helix.
LMNT_CREATE_VOICECreates a custom voice in LMNT by training on uploaded audio samples. The voice can then be used for text-to-speech synthesis. Returns the voice ID and metadata upon successful creation. The voice may be in 'training' state initially before becoming 'ready'.
Input parameters
The display name for this voice.
The type of voice to create. Use 'instant' for quick voice cloning or 'professional' for higher quality. Defaults to 'instant'.
Audio file to train the voice. Supported formats: WAV, MP3, MP4, M4A, WebM. Maximum file size: 250 MB.
Optional gender tag for this voice (e.g., 'male', 'female', 'nonbinary'). This is metadata only and does not affect voice generation.
Whether to apply audio processing to reduce background noise. Set to true for unclean audio, but note this may degrade quality in some cases. Defaults to false.
Optional text description of this voice for organizational purposes.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_DELETE_VOICE_INFODeletes a voice from your LMNT account. This operation permanently removes the voice and cancels any pending operations on it. This action cannot be undone. Only voices owned by you (owner='me') can be deleted; system voices cannot be deleted. Use case: Remove custom voices that are no longer needed to manage your voice library.
Input parameters
The unique identifier of the voice to delete. This must be a voice owned by you (voices with owner='me'). System voices cannot be deleted.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_GENERATE_SPEECH_WITH_METADATAGenerates speech from text and returns JSON with base64-encoded audio and optional word-level timing metadata. Use when you need the synthesis seed or word timestamps for subtitle synchronization. For lower latency without metadata, use the Synthesize Speech action instead.
Input parameters
Integer seed for reproducible speech variations. Use the same seed to replicate a specific output.
The text to synthesize into speech (max 5000 characters including spaces).
When true, saves the synthesis clip to your clip library for debugging purposes.
The synthesis model to use (default: 'blizzard').
Controls speech stability (0-1 range, default: 0.8). Lower values produce more consistent speech.
The voice ID to use for speech synthesis (e.g., 'lily', 'leah', 'daniel'). Use the List Voices action to get available voice IDs.
Output audio format. Streamable formats (generate faster): mp3 (default), ulaw, webm, pcm_s16le, pcm_f32le. Non-streamable: aac, wav.
ISO 639-1 two-letter language code (e.g., 'en', 'es', 'fr'). Auto-detected by default, but specifying language improves generation speed.
Audio sample rate in Hz. Options: 8000, 16000, or 24000.
Controls speech expressiveness (numeric value, default: 1.0). Higher values increase variation.
When true, returns word-level duration timestamps in the response. Useful for synchronizing subtitles or animations with speech.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_GET_ACCOUNTRetrieves account information including subscription plan details and current usage statistics.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_GET_VOICE_INFOGets metadata for a specific LMNT voice, including active status, supported languages, and plan availability. Useful for validating a voice ID before using it in synthesis requests.
Input parameters
The ID of the voice to retrieve. Can be obtained from the list voices endpoint.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_GET_VOICES_LISTRetrieves a list of available voices from LMNT. Returns both system-provided preset voices and any custom voices you have created. Use filters to narrow results by ownership (system vs custom) or starred status. Each voice includes details like ID, name, description, gender, state, and preview URL.
Input parameters
Filter voices by owner. Options: 'all' (default, shows all voices), 'system' (only LMNT preset voices), 'me' (only your custom voices).
Filter to show only starred voices. Set to true to return only voices you have starred, false (default) to show all voices.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_SYNTHESIZE_SPEECHSynthesizes speech from text using LMNT's AI voices. Converts text (up to 5000 characters) into natural-sounding speech audio using a specified voice. Returns base64-encoded audio at `data.response_data.audio` — decode before saving or passing to other tools. Supports multiple audio formats and quality settings for different use cases.
Input parameters
Integer seed for reproducible speech variations. Use the same seed to replicate a specific output.
The text to synthesize into speech (max 5000 characters including spaces). For texts exceeding 5000 characters, split into chunks and call separately, keeping `voice`, `model`, `format`, and `sample_rate` identical across all chunks to avoid audible seams.
When true, saves the synthesis clip to your clip library for debugging purposes.
The synthesis model to use (default: 'blizzard').
Controls speech stability (0-1 range, default: 0.8). Lower values produce more consistent speech.
The voice ID to use for speech synthesis (e.g., 'lily', 'leah', 'daniel'). Use the List Voices action to get available voice IDs.
Output audio format. Streamable formats (generate faster): mp3 (default), ulaw, webm, pcm_s16le, pcm_f32le. Non-streamable: aac, wav.
ISO 639-1 two-letter language code (e.g., 'en', 'es', 'fr'). Auto-detected by default, but specifying language improves generation speed.
Audio sample rate in Hz. Options: 8000, 16000, or 24000 (default).
Controls speech expressiveness (numeric value, default: 1.0). Higher values increase variation.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
LMNT_UPDATE_VOICEUpdates information about a specific voice in LMNT. You can update the name, description, gender, starred status, and unfreeze state of a voice. Note: Only user-owned voices (owner='me') can have their name, description, and gender updated. System voices can only be starred/unstarred.
Input parameters
The unique identifier of the voice to update
The display name for this voice.
A tag describing the gender of this voice, e.g. male, female, nonbinary.
If true, adds this voice to your starred list. If false, removes it from your starred list.
If true, unfreezes this voice and upgrades it to the latest model.
A description of this voice.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
No publicly available marketplace agent is found using this tool yet. There are 33 agents privately built on Nagent that already use LMNT.
Build on Nagent
Connect LMNT to any Nagent agent in minutes — no API key management, no boilerplate. Just configure and deploy.
The five questions agent builders ask before adopting a new integration.
Open the External Integrations panel inside Nagent (app.nagent.ai/externalIntegration), find LMNT, and click "Connect Now." You'll authenticate with an API key — Nagent handles credential storage and refresh automatically. Once connected, LMNT is available to any agent in your workspace.
No. Nagent provides no-code integration for every tool. Once LMNT is connected, you configure its 8 actions directly in the agent builder UI — no API calls, no boilerplate, no schema management.
Helix — Nagent's agentic agent builder — lets you drop LMNT steps into any workflow visually. Pick an action (e.g., one of those listed above), fill in the inputs (Helix knows the required vs. optional schema for each parameter), and connect it to upstream/downstream steps. Triggers run as the entry point of an agent, so when a LMNT event fires, the agent kicks off automatically.
Every LMNT action and trigger ships with a fully-typed schema — input parameters with name, type, required flag, and description, plus the output payload shape. The schemas are documented in the sections above. Helix uses these schemas to validate your configuration at build time and to type-check the data flowing between steps.
Yes. While LMNT ships with 8 pre-built artificial intelligence actions, you can layer custom logic around them inside Helix — pre/post-processing steps, conditional branches, retries, or stitching LMNT together with other connected tools. For deeper customization, talk to our team about Nagent's Agentic AI Lab — forward-deployed engineers who build LMNT-based workflows tailored to your business.