Complete suite of 13 nodes providing full access to ElevenLabs AI audio capabilities including text-to-speech, voice cloning, dubbing, sound effects, music generation, and more.

📋 Table of Contents

Features
Installation
Getting Your API Key
Available Nodes
Quick Start
Troubleshooting
License

✨ Features

13 Specialized Nodes covering all ElevenLabs API endpoints
Auto-Refresh voices when API key changes
V3 Model Support with emotional expressiveness (70+ languages)
Voice Cloning from audio samples
Dubbing for video localization
Sound Effects generation from text
Music Generation with AI
Voice Isolation to remove background noise
Account Management to check credits and usage
Comprehensive Error Logging for debugging

🚀 Installation

Method 1: ComfyUI Manager (Recommended)

Open ComfyUI Manager
Search for "Elevenlabs-ComfyUI"
Click Install

Method 2: Manual Installation

cd  ComfyUI/custom_nodes/
git  clone  https://github.com/karthikg-09/Elevenlabs-ComfyUI.git
cd  Elevenlabs-ComfyUI
pip  install  -r  requirements.txt

Method 3: Portable/Standalone Installation

cd  ComfyUI_windows_portable/ComfyUI/custom_nodes/
git  clone  https://github.com/karthikg-09/Elevenlabs-ComfyUI.git
cd  Elevenlabs-ComfyUI
../../python_embeded/python.exe  -m  pip  install  -r  requirements.txt

Restart ComfyUI after installation.

🔑 Getting Your API Key

Visit ElevenLabs
Sign up for a free account (or log in)
Navigate to your Profile Settings
Click "Generate API Key"
Copy the API key
Paste it into the api_key field in any node

Note: Keep your API key secure and never share it publicly.

📦 Available Nodes

All nodes are located in the "ElevenLabs" category in ComfyUI's Add Node menu.

1. ElevenLabs TTS (Text-to-Speech)

Convert text to natural-sounding speech with full control over voice characteristics.

Inputs:

api_key (STRING) - Your ElevenLabs API key
text (STRING) - Text to synthesize (3K-40K character limit depending on model)
voice (DROPDOWN) - Voice selection from your account (auto-populated)
model (DROPDOWN) - AI model (eleven_v3, eleven_multilingual_v2, eleven_turbo_v2_5, etc.)
stability (FLOAT 0-1) - Voice consistency - Lower = more expressive, Higher = more stable
similarity_boost (FLOAT 0-1) - How closely output matches original voice characteristics
style (FLOAT 0-1) - Emotional expressiveness - Higher = more dramatic (especially for eleven_v3)
use_speaker_boost (BOOLEAN) - Enhance voice clarity and cloning accuracy (recommended: ON)
language_code (DROPDOWN) - Target language (or "auto" for automatic detection)
output_format (DROPDOWN) - Audio format: mp3_44100_128, pcm_16000, pcm_44100, etc.
seed (INT) - Seed for reproducible generation (-1 for random)

Outputs:

AUDIO - Generated speech audio (compatible with ComfyUI audio nodes)

Use Cases: Voiceovers, audiobooks, narration, character dialogue, e-learning content

2. ElevenLabs Speech-to-Text

Transcribe spoken audio into text with high accuracy (99 languages supported).

Inputs:

api_key (STRING) - Your ElevenLabs API key
audio (AUDIO) - Audio input to transcribe
model (DROPDOWN) - Transcription model (scribe_v1 or scribe_v1_experimental)
language (DROPDOWN) - Target language for transcription (or "auto" for detection)

Outputs:

transcription (STRING) - Text transcription of the audio

Use Cases: Subtitles, transcription services, voice-to-text workflows

3. ElevenLabs Sound Effects

Generate cinematic sound effects from text descriptions.

Inputs:

api_key (STRING) - Your ElevenLabs API key
text (STRING) - Description of the sound effect (e.g., "thunder rumbling", "door creaking")
duration_seconds (FLOAT 0.5-22) - Length of the generated sound effect
prompt_influence (FLOAT 0-1) - How closely to follow the prompt - Higher = more literal

Outputs:

AUDIO - Generated sound effect

Use Cases: Game audio, film/video sound design, podcast effects, ambient sounds

4. ElevenLabs Voice Isolator

Remove background noise and isolate clean voice audio from recordings.

Inputs:

api_key (STRING) - Your ElevenLabs API key
audio (AUDIO) - Audio input with background noise

Outputs:

AUDIO - Clean audio with voice isolated

Use Cases: Podcast cleanup, interview enhancement, noise reduction, vocal extraction

5. ElevenLabs Voice Changer (Speech-to-Speech)

Transform the voice in an audio recording to a different voice while preserving speech patterns.

Inputs:

api_key (STRING) - Your ElevenLabs API key
audio (AUDIO) - Original audio with voice to transform
target_voice (DROPDOWN) - Voice to transform into (from your account)
model (DROPDOWN) - STS model (eleven_english_sts_v2 or eleven_multilingual_sts_v2)
stability (FLOAT 0-1) - Voice consistency in transformation
similarity_boost (FLOAT 0-1) - How closely to match target voice
style (FLOAT 0-1) - Expressiveness of transformed voice
use_speaker_boost (BOOLEAN) - Enhance transformation accuracy

Outputs:

AUDIO - Audio with transformed voice

Use Cases: Voice anonymization, character voice acting, voice modification, creative projects

6. ElevenLabs Dubbing

Automatically dub videos or audio into different languages while preserving voice characteristics.

Inputs:

api_key (STRING) - Your ElevenLabs API key
audio (AUDIO) - Source audio to dub
target_language (DROPDOWN) - Language to translate/dub into
source_language (DROPDOWN) - Original language (or "auto" for detection)
num_speakers (INT 1-10) - Number of distinct speakers in the audio
watermark (BOOLEAN) - Add ElevenLabs watermark (required for free tier)

Outputs:

AUDIO - Dubbed audio in target language

Use Cases: Video localization, multilingual content, international distribution

7. ElevenLabs Voice Design

Create custom voices from text descriptions using AI.

Inputs:

api_key (STRING) - Your ElevenLabs API key
text (STRING) - Text description of desired voice (e.g., "young female, British accent, calm")
gender (DROPDOWN) - Voice gender: male, female, or neutral
age (DROPDOWN) - Voice age: young, middle_aged, or old
accent (DROPDOWN) - Voice accent: american, british, australian, etc.
accent_strength (FLOAT 0-2) - Strength of the accent - 1.0 is neutral

Outputs:

voice_id (STRING) - ID of the generated voice
AUDIO - Preview of the generated voice

Use Cases: Rapid character voice creation, A/B testing voices, prototyping

8. ElevenLabs Voice Clone

Clone a voice from audio samples for use in text-to-speech.

Inputs:

api_key (STRING) - Your ElevenLabs API key
voice_name (STRING) - Name for the cloned voice
audio (AUDIO) - Audio sample of the voice to clone (1-5 minutes recommended)
description (STRING) - Optional description of the voice

Outputs:

voice_id (STRING) - ID of the cloned voice (use in TTS node)
status (STRING) - Status message

Use Cases: Personal voice cloning, character voice creation, voice preservation

Note: Requires appropriate subscription tier. Only clone voices you have permission to use.

9. ElevenLabs Music Generation

Generate studio-quality music from text prompts.

Inputs:

api_key (STRING) - Your ElevenLabs API key
prompt (STRING) - Description of desired music (e.g., "upbeat electronic dance music")
duration_seconds (FLOAT 1-300) - Length of generated music
model (DROPDOWN) - Music model (eleven_music_gen_v1)
temperature (FLOAT 0.1-2.0) - Creativity/randomness - Higher = more variation

Outputs:

AUDIO - Generated music

Use Cases: Background music, game soundtracks, video scoring, creative projects

Note: Requires specific subscription tier with music generation access.

10. ElevenLabs User Info

Check your account status, credits, character usage, and subscription details.

Inputs:

api_key (STRING) - Your ElevenLabs API key
refresh (BOOLEAN) - Force refresh account data from API

Outputs:

user_info (STRING) - Formatted account information including:
Email and subscription tier
Character usage and limits
Remaining character quota
Voice cloning slots
API access status

Use Cases: Monitor usage, check remaining credits, verify subscription status

11. ElevenLabs Refresh Voices

Manually refresh the cached list of voices and models from your account.

Inputs:

api_key (STRING) - Your ElevenLabs API key
refresh_voices (BOOLEAN) - Refresh voice list from API
refresh_models (BOOLEAN) - Refresh available models list

Outputs:

status (STRING) - Refresh status message

Use Cases: Update voice list after cloning, force cache refresh, sync account changes

Note: TTS and Voice Changer nodes automatically refresh when API key changes. This node is for manual control.

12. ElevenLabs History

View your generation history and retrieve previous audio generations.

Inputs:

api_key (STRING) - Your ElevenLabs API key
page_size (INT 1-100) - Number of history items to retrieve

Outputs:

history (STRING) - Formatted list of recent generations with metadata

Use Cases: Track usage, review previous generations, audit API calls

13. ElevenLabs Voice Manager

Manage voices in your account (list, delete, or get details).

Inputs:

api_key (STRING) - Your ElevenLabs API key
action (DROPDOWN) - Action to perform: list, get_details, or delete
voice_id (STRING) - Voice ID (required for get_details and delete actions)

Outputs:

result (STRING) - Result of the action

Use Cases: Voice library management, cleanup unused voices, voice information retrieval

🎯 Quick Start

Basic Text-to-Speech Workflow

Add Node: Right-click → Add Node → ElevenLabs → ElevenLabs TTS
Enter API Key: Paste your ElevenLabs API key
Enter Text: Type or connect text input
Select Voice: Choose from dropdown (auto-populated from your account)
Select Model: Choose eleven_v3 for emotions, or eleven_turbo_v2_5 for speed
Adjust Settings:

stability: 0.5 (default)
similarity_boost: 0.75 (default)
style: 0.7 for emotional delivery, 0.0 for neutral

Run Workflow: Queue Prompt
First Run: Voices refresh automatically
Reload Node: Right-click node → Reload Node to see updated voice list

Voice Cloning Workflow

Add ElevenLabs Voice Clone node
Enter API key and voice name
Connect audio sample (1-5 minutes of clear speech)
Run to create cloned voice
Copy the voice_id output
Use in ElevenLabs TTS node for generation

Emotional Speech (V3 Model)

Add ElevenLabs TTS node
Select eleven_v3 model
Set style to 0.7-1.0
Set stability to 0.3-0.5
Use descriptive text with emotional context
Generate dramatic, expressive speech in 70+ languages

🎛️ Parameter Guide

Voice Settings

Stability (0.0-1.0)
Low (0.0-0.3) = Expressive, variable speech
Mid (0.4-0.6) = Balanced delivery
High (0.7-1.0) = Consistent, predictable
Similarity Boost (0.0-1.0)
Low (0.0-0.3) = Generic voice qualities
Mid (0.4-0.6) = Match original voice
High (0.7-1.0) = Exact voice match
Style (0.0-1.0)
Low (0.0-0.3) = Neutral, calm tone
Mid (0.4-0.6) = Moderate emotion
High (0.7-1.0) = Dramatic, intense delivery

Model Selection

eleven_v3
Latency: High | Languages: 70+ | Char Limit: 3,000
Best for: Emotional content, audiobooks, dramatic delivery
eleven_multilingual_v2
Latency: Medium | Languages: 29 | Char Limit: 10,000
Best for: General purpose, high quality speech
eleven_turbo_v2_5
Latency: Low | Languages: 32 | Char Limit: 40,000
Best for: Speed + quality balance, long content
eleven_flash_v2_5
Latency: Ultra-low (~75ms) | Languages: 32 | Char Limit: 40,000
Best for: Real-time applications, voice agents

For emotional speech: Use eleven_v3 with high style (0.7-1.0)

For fast generation: Use eleven_flash_v2_5 or eleven_turbo_v2_5

For long text: Use eleven_turbo_v2_5 (40K char limit)

🔧 Troubleshooting

Voices Not Appearing

Cause: Voice list needs refresh after adding new voices.

Solution:

Enter/modify API key in TTS node
Run the node once (Queue Prompt)
Console shows: 🔄 New API key detected - Auto-refreshing voices...
Right-click node → Reload Node (or F5)
Voices appear in dropdown

Alternative: Use ElevenLabs Refresh Voices node

Error: Invalid API Key

Cause: API key is incorrect, expired, or not properly formatted.

Solution:

Verify API key at ElevenLabs Settings
Generate new API key if needed
Copy entire key including any dashes or special characters
Ensure no extra spaces before/after key

Error: Character Limit Exceeded

Cause: Text exceeds model's character limit.

Solution:

eleven_v3: Max 3,000 characters
eleven_multilingual_v2: Max 10,000 characters
eleven_turbo_v2_5: Max 40,000 characters
Split long text into chunks or use a model with higher limit

Error: Insufficient Credits

Cause: Account has run out of character quota.

Solution:

Use ElevenLabs User Info node to check remaining credits
Upgrade subscription at ElevenLabs Pricing
Wait for monthly quota reset

Audio Quality Issues

Common Fixes:

Increase stability for more consistent output
Increase similarity_boost for better voice matching
Enable use_speaker_boost for cloned voices
Try different voices (some are more stable)
Use higher quality output_format (e.g., pcm_44100)

Console Errors and Logging

All nodes output detailed error messages to the ComfyUI console/terminal:

API request details (voice ID, model, settings)
HTTP status codes and error responses
Helpful suggestions for fixing issues

Always check the console for debugging information.

📚 Additional Resources

ElevenLabs Documentation: elevenlabs.io/docs
API Reference: elevenlabs.io/docs/api-reference
Model Information: elevenlabs.io/docs/models
Voice Library: elevenlabs.io/voice-library
Pricing: elevenlabs.io/pricing

🤝 Contributing

Contributions are welcome! Please:

Fork the repository
Create a feature branch
Make your changes
Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

⚠️ Important Notes

API Key Security: Never share your API key publicly or commit it to version control
Voice Cloning Ethics: Only clone voices you have permission to use
Rate Limits: Respect ElevenLabs API rate limits (varies by subscription tier)
Terms of Service: Use in compliance with ElevenLabs Terms of Service
Data Privacy: Review ElevenLabs Privacy Policy for data handling

🎉 Credits

Created and maintained by the karthikg-09.

Built for the ComfyUI community.

Enjoy creating amazing AI audio with ElevenLabs in ComfyUI! 🎙️✨