Eleven Labs Skill

AI-powered voice generation, voice cloning, and sound effects using Eleven Labs.

Getting API Access

1. Sign up at https://elevenlabs.io
2. Go to Profile → API Keys
3. Copy your API key

Pricing: Free tier includes limited characters/month. Paid plans for more usage and voice cloning.

Setup

# Install dependencies
pip3 install elevenlabs

# Configure API key
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py setup YOUR_API_KEY

Commands

List Available Voices

python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py voices
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py voices --category premade

Generate Speech (Text-to-Speech)

# Basic usage (default voice)
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py speak "Hello, welcome to our film."

# With specific voice
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py speak "Hello world" --voice "Rachel"
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py speak "Hello world" --voice "Adam"

# With specific model
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py speak "Hello" --model eleven_multilingual_v2

Output saved to ~/.claude/skills/eleven-labs-skill/output/

Clone a Voice

# Provide 1-3 audio samples of the voice to clone
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py clone "My Voice" sample1.mp3 sample2.mp3

# With description
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py clone "Narrator" audio.mp3 --description "Deep narrator voice"

Tips for voice cloning:

• Use clean audio with minimal background noise
• 1-3 minutes of audio works best
• Consistent speaking style improves results

Generate Sound Effects

python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py sfx "thunder rolling in the distance"
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py sfx "footsteps on gravel" --duration 3
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py sfx "crowd cheering in a stadium" --duration 10

List Models

python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py models

View Generation History

python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py history
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py history --limit 50

Delete a Cloned Voice

python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py delete-voice VOICE_ID

Output

All audio files saved to ~/.claude/skills/eleven-labs-skill/output/

{
  "status": "success",
  "file": "/Users/you/.claude/skills/eleven-labs-skill/output/speech_20260202_153000.mp3",
  "text": "Hello, welcome to our film.",
  "voice": "21m00Tcm4TlvDq8ikWAM"
}

Popular Voices

Name	Style	Good For
Rachel	Neutral American	Narration, general
Adam	Deep American	Trailers, dramatic
Bella	British	Storytelling
Antoni	American	Conversational
Elli	American Young	Animation, youth
Josh	American Deep	Documentary
Arnold	American Crisp	Professional
Sam	American Raspy	Character work

Use with Film-Maker Skill

This skill integrates with the film-maker skill for full production:

# Generate dialogue for a scene
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py speak "To be or not to be" --voice "Josh"

# Generate ambient sound
python3 ~/.claude/skills/eleven-labs-skill/eleven_labs_skill.py sfx "rain on window" --duration 30

Requirements

elevenlabs>=1.0.0

Security Notes

• API key stored in ~/.claude/skills/eleven-labs-skill/config.json
• Can also use ELEVENLABS_API_KEY environment variable

#elevenlabs #voice #tts #audio #film