Skip to content
AgentHosted agent

Cassette AI

Generative music, sound effect and speech models via one API for games and creator apps

Cassette AI is a paid hosted agent for music. It is available on API and Web.

Pricing
Paid
Platforms
API, Web
Listing updated

About Cassette AI

The site describes three generative audio models: one for music, one for sound effects and one for text-to-speech. Music output can be a short sample or a full track of up to three minutes at 44.1 kHz stereo, while sound effects are capped at 30 seconds and are pitched at real-time use cases like games and AR/VR.

Access is through a single API shape, with the page showing calls via the fal.subscribe() client from JavaScript, Python and cURL. Switching between the music, sound effect and speech endpoints only changes the model ID, which keeps integration consistent across modalities.

Billing is metered per output second rather than by subscription seat. The page lists a per-minute rate for music and a per-generation rate for sound effects, and says dedicated capacity or on-device licensing is handled by sales contact.

What it does

Music generation up to 3 minutes
Creates tracks from text prompts, including tempo and key, at 44.1 kHz stereo in WAV.
Short sound effects
Renders effects up to 30 seconds from a text description, with outputs designed to loop.
Voice synthesis with streaming
Text-to-speech with a streaming-first API and optional zero-shot voice cloning from a short sample.
Deterministic seeds
Music and sound effect outputs can be reproduced using a fixed seed, according to the site.
Unified API shape
The same call structure works across the music, sound effect and speech models.
Metered pricing
Pay per generated audio second with no monthly commitment or seat fees.

Good for

  • Generating background music that adapts while a player is in a game
  • Producing sound effects on demand for game or video production pipelines
  • Adding synthetic voice narration to creator apps or real-time applications
  • Creating loopable effects for AR/VR experiences

Pricing

Paid

Pay-per-use: $0.02 per output minute for music, $0.01 per sound effect generation; enterprise pricing via sales.

Prices change. Check cassetteai.com before you buy.

Cassette AI FAQ

What is the maximum length of a generated music track?

The site says full tracks go up to 3 minutes per call, with a 10 to 180 second duration range listed.

How is usage billed?

Billing is metered per output second with no monthly commit or developer seats. The page lists $0.02 per output minute for music and $0.01 per sound effect generation.

How do I call the models?

The page shows examples using the fal.subscribe() client with the CassetteAI music and sound effects model IDs, and also shows cURL usage.

Is Cassette AI free?

No. Cassette AI is a paid product.

Where this comes from

  • Description: written from the project's website.