johnalbertini14-glitch Description
AudioMind is a comprehensive, zero-configuration audio generation skill that acts as a smart dispatcher for all your audio needs. It intelligently routes your natural language requests to a vast library of over 18 specialized audio models from leading providers like ElevenLabs and fal.ai.
Simply install this skill, and your agent instantly gains the ability to perform:
- Text-to-Speech (TTS): Using 9 different models for various languages, styles, and latencies.
- Music Generation: Creating royalty-free instrumental tracks in multiple genres.
- Sound Effect (SFX) Generation: Producing a wide range of sound effects from text descriptions.
- Voice Cloning: Cloning a voice from an audio sample to generate new speech.
This skill is powered by a backend proxy service that securely manages API keys and usage, providing a seamless, plug-and-play experience for the end-user.
How It Works
- Installation: The user installs the
audiomindskill. - Request: The user makes a natural language request for audio (e.g., "Create a sound effect of a passing train").
- Smart Routing: AudioMind analyzes the request and consults its internal model registry to select the best model for the job.
- Proxy Call: The skill sends a structured request to the AudioMind Proxy Service.
- API Execution: The proxy service calls the appropriate underlying API (ElevenLabs or fal.ai) with the correct parameters.
- Result: The generated audio is returned to the user.
Usage
This skill is designed to be used with natural language. The smartRoute logic will automatically determine the best model. However, you can also specify a model directly.
Smart Routing (Recommended)
# TTS
"Narrate this: Hello, world!"
# Music
"Compose a 30-second upbeat lo-fi hip hop track."
# SFX
"I need the sound of a thunderstorm."
# Voice Clone
"Clone the voice from this audio file and say: Welcome to the future."
Direct Model Selection (Advanced)
{
"action": "tts",
"text": "This is a test.",
"model": "minimax-tts-hd"
}
Model Registry (v2.0.0)
AudioMind v2.0 intelligently routes requests to the following models:
| Type | Model ID | Provider | Description |
|---|---|---|---|
| TTS | elevenlabs-tts-v3 |
ElevenLabs | Most expressive, multilingual |
elevenlabs-tts-v2 |
ElevenLabs | Stable, 29 languages | |
elevenlabs-tts-turbo |
ElevenLabs | Ultra-low latency, 32 languages | |
minimax-tts-hd |
fal.ai | High quality, multilingual | |
minimax-tts-2.6-hd |
fal.ai | MiniMax Speech-2.6 HD | |
minimax-tts-2.8-hd |
fal.ai | Latest MiniMax TTS | |
minimax-tts-2.8-turbo |
fal.ai | Fast, low latency | |
chatterbox-tts |
fal.ai | Emotion-aware TTS | |
playai-dialog |
fal.ai | Multi-speaker dialogue generation | |
| Voice Clone | dia-voice-clone |
fal.ai | Clone any voice from an audio sample |
| Music | elevenlabs-music |
ElevenLabs | Composition-plan based music |
beatoven-music |
fal.ai | Royalty-free instrumental, 10+ genres | |
cassetteai-music |
fal.ai | Fast music generation | |
| SFX | elevenlabs-sfx |
ElevenLabs | Text-to-sound-effects |
beatoven-sfx |
fal.ai | Professional sound effects | |
mirelo-video-to-audio |
fal.ai | Generate synced audio for video | |
mirelo-video-to-video |
fal.ai | Add sound track to video |
Commercial Use
This skill includes a free tier of 100 generations. For unlimited use, please upgrade to AudioMind Pro by visiting our Gumroad page (link will be provided when the free limit is reached).
To activate Pro, set the AUDIOMIND_PRO_KEY environment variable with the key you receive after purchase.