🧪 Skills

Elevenlabs Pro

--- name: elevenlabs-pro description: ElevenLabs advanced TTS for converting text to speech, listing voices, and managing credits license: MIT metadata: version: 1.0.0 author: Jack2 tags: tts, a

v0.1.0
❤️ 0
⬇️ 254
👁 1
Share

Description


name: elevenlabs-pro description: ElevenLabs advanced TTS for converting text to speech, listing voices, and managing credits license: MIT metadata: version: 1.0.0 author: Jack2 tags: tts, audio, elevenlabs, voice

ElevenLabs Pro Skill

Advanced text-to-speech using the ElevenLabs API. Supports voice selection, language/gender filtering, audio file generation, and credit tracking.

Prerequisites

Set your API key as an environment variable:

export ELEVENLABS_API_KEY=sk_...

Or pass it directly with --api-key.

Usage

Convert text to speech

python3 scripts/elevenlabs.py "Hello world" --voice Sarah --output audio.mp3

List available voices

python3 scripts/elevenlabs.py --list-voices
python3 scripts/elevenlabs.py --list-voices --lang en --gender female
python3 scripts/elevenlabs.py --list-voices --json

Check remaining credits

python3 scripts/elevenlabs.py --credits

Options

Option Default Description
--voice NAME Rachel Voice name (partial match supported)
--voice-id ID Direct voice ID (overrides --voice)
--output PATH output.mp3 Output MP3 file path
--model ID eleven_turbo_v2_5 Model ID
--stability 0.5 Voice stability (0–1)
--similarity 0.75 Similarity boost (0–1)
--style 0.0 Style exaggeration (0–1)
--lang CODE Filter voices by language (e.g. en, fr)
--gender Filter voices by gender (male/female)
--json Output as JSON
--api-key KEY API key (overrides env var)

Available Models

Model ID Description
eleven_turbo_v2_5 Fastest, lowest latency (free tier supported)
eleven_multilingual_v2 Best quality, multilingual
eleven_flash_v2_5 Ultra-low latency

Importable API

import sys
sys.path.insert(0, "path/to/skills/elevenlabs-pro/scripts")
from elevenlabs import list_voices, get_voice_id, text_to_speech, get_credits

api_key = "sk_..."
# TTS
voice_id = get_voice_id(api_key, "Sarah")
path = text_to_speech(api_key, "Hello!", voice_id, "out.mp3")

# Credits
info = get_credits(api_key)
print(info["characters_remaining"])

References

See references/voices.md for popular voices and voice parameter guidance.

Reviews (0)

Sign in to write a review.

No reviews yet. Be the first to review!

Comments (0)

Sign in to join the discussion.

No comments yet. Be the first to share your thoughts!

Compatible Platforms

Pricing

Free

Related Configs