Audio Generation

TF Design supports audio generation through the ToothFairyAI backend. Three audio categories are available: music, voice, and sound effects.

Available Models

Model	Alias	Category	Output
Mystica Music	mystica_music	Music	Instrumental tracks, background music
Mystica Voice	mystica_voice	Voice	Text-to-speech narration, voiceovers
Mystica SFX	mystica_sfx	Sound Effects	Foley, ambient sounds, UI sounds

How It Works

1. Provide a prompt describing the desired audio.
2. The Desktop app sends the request to the ToothFairyAI backend.
3. The model generates audio and returns it as a file.
4. The output appears as a download chip in the chat.
5. Audio can be embedded in artifacts or downloaded.

Use Cases

Music

• Background music for presentations and decks.
• Ambient tracks for interactive prototypes.
• Brand audio signatures and jingles.

Prompt example: "Upbeat electronic background music, 120 BPM, 30 seconds, suitable for a product demo video."

Voice

• Narration for slide decks and video explainers.
• Voiceovers for landing page hero sections.
• Accessibility audio descriptions.

Prompt example: "Professional female voice, British English, reading: 'Welcome to Insight, the analytics platform for data teams.'"

Sound Effects

• UI interaction sounds (clicks, hovers, transitions).
• Ambient environment sounds.
• Foley for video content.

Prompt example: "Soft digital click sound, short, suitable for a button press interaction."

Using Generated Audio in Artifacts

Generated audio files are saved to the project workspace and can be:

• Embedded in HTML — <audio src="assets/track.mp3" controls> or via JavaScript Audio() .
• Triggered on interaction — play sounds on button clicks or transitions.
• Downloaded — click the download chip to save locally.
• Included in exports — audio files are part of the ZIP archive.

Internal Endpoints

The following endpoints are exposed by the running Desktop app's embedded runtime for in-app tools and the agent:

Endpoint	Method	Description
/api/projects/:id/media/generate	POST	Generate audio (specify audio model)
/api/projects/:id/media/tasks	GET	List media tasks
/api/media/tasks/:id/wait	GET	Wait synchronously for a task to finish