Skip to content
Model infoModelsSnapshot: 15 Oct 2026

CassetteAI / CassetteAI audio generators

CassetteAI Sound Effects Generator

Text-guided sound effects with 1–30-second WAV output through fal, billed per generation.

availableClosed weightsSound effects generation
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
13 of 29 specification fields collected
Audio output durationInteger 1–30 secondsSee specification notes below
Audio formatsWAVSee specification notes below
Input billingSee ratesfal API
Output billingSee ratesPublished units and conditions below

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
Not collected
Maximum input tokens
Not collected
Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected

JSON file envelopes do not establish constrained structured text generation.

Streaming
Not collected

Streaming is advertised, but a model-specific streaming transport contract has not been verified.

Batch processing
Not collected

Queueing multiple requests is not a documented discounted Batch API.

Fine-tuning
Not collected

Dedicated capacity and on-device licensing do not establish model fine-tuning support.

Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Not collected
Languages
Not collected
Context caching
Not collected

No model-specific prompt caching eligibility, controls, lifetime or rates documented.

Output duration
Integer 1–30 secondsSource for Output duration (opens in a new tab)

Required duration; no default. Applies to this fal contract, not every edge deployment.

Output formats
WAVSource for Output formats (opens in a new tab)

Returns a signed audio_file URL. Generic File PNG examples do not describe this operation.

Audio sample rates
44.1 kHzSource for Audio sample rates (opens in a new tab)

Provider-reported audio format; no inference request used for measurement.

Audio bit depth
16-bitSource for Audio bit depth (opens in a new tab)

Homepage output-quality claim; the fal schema does not expose an adjustable bit-depth control.

Required request fields
prompt string and integer duration. No prompt character/token maximum, seed, reference audio, negative prompt or streaming field is exposed in this schema.Source for Required request fields (opens in a new tab)
Output transport
One audio_file object with a signed WAV URL; expiry time is not documented by this model schema.Source for Output transport (opens in a new tab)
Streaming claim and API scope
Provider marketing describes real-time streaming and thirty-second chunks. The reviewed fal contract exposes queued file output, without a streaming-audio route.Source for Streaming claim and API scope (opens in a new tab)
Random seed
Not collected

Homepage describes deterministic seeds, but the public fal schema has no seed control. A seed API is not inferred.

Published price
USD 0.01 per generation on falSource for Published price (opens in a new tab)
Deployment differences
The edge homepage card describes 100 ms–12-second SFX, while its hosted pricing section and fal schema allow 1–30 seconds. Keep those deployment contracts separate.Source for Deployment differences (opens in a new tab)

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

API model ID
cassetteai/sound-effects-generator
Access
POST https://queue.fal.run/cassetteai/sound-effects-generator with Authorization: Key <FAL_KEY>; retrieve status/result using returned URLs, PUT cancellation. Current playground and pricing are publicly listed; no paid generation was performed.
Regions
Hosted API; inference residency unverified
Context / input limit
Not collected
Maximum input tokens
Not collected
Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected

JSON file envelopes do not establish constrained structured text generation.

Streaming
Not collected

Streaming is advertised, but a model-specific streaming transport contract has not been verified.

Batch processing
Not collected

Queueing multiple requests is not a documented discounted Batch API.

Fine-tuning
Not collected

Dedicated capacity and on-device licensing do not establish model fine-tuning support.

Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Not collected
Languages
Not collected
Context caching
Not collected

No model-specific prompt caching eligibility, controls, lifetime or rates documented.

Output duration
Integer 1–30 secondsSource for Output duration (opens in a new tab)

Required duration; no default. Applies to this fal contract, not every edge deployment.

Output formats
WAVSource for Output formats (opens in a new tab)

Returns a signed audio_file URL. Generic File PNG examples do not describe this operation.

Audio sample rates
44.1 kHzSource for Audio sample rates (opens in a new tab)

Provider-reported audio format; no inference request used for measurement.

Audio bit depth
16-bitSource for Audio bit depth (opens in a new tab)

Homepage output-quality claim; the fal schema does not expose an adjustable bit-depth control.

Required request fields
prompt string and integer duration. No prompt character/token maximum, seed, reference audio, negative prompt or streaming field is exposed in this schema.Source for Required request fields (opens in a new tab)
Output transport
One audio_file object with a signed WAV URL; expiry time is not documented by this model schema.Source for Output transport (opens in a new tab)
Streaming claim and API scope
Provider marketing describes real-time streaming and thirty-second chunks. The reviewed fal contract exposes queued file output, without a streaming-audio route.Source for Streaming claim and API scope (opens in a new tab)
Random seed
Not collected

Homepage describes deterministic seeds, but the public fal schema has no seed control. A seed API is not inferred.

Published price
USD 0.01 per generation on falSource for Published price (opens in a new tab)
Deployment differences
The edge homepage card describes 100 ms–12-second SFX, while its hosted pricing section and fal schema allow 1–30 seconds. Keep those deployment contracts separate.Source for Deployment differences (opens in a new tab)
Cancellation and task deletion
Queued requests are removed on cancellation; in-progress cancellation is a runner signal and may not stop execution.Source for Cancellation and task deletion (opens in a new tab)
Concurrent requests
Account-wide concurrency starts at two and self-serve limits scale up to forty based on paid invoices; high-demand endpoints and enterprise contracts can have separate limits. Check account-specific capacity.Source for Concurrent requests (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 14 Oct 2026
CategoryUsageRateUnit
Sound effectsGenerated sound effect1$0.01generation

Rate conditions

  1. Flat charge per generation, within the 1–30-second duration contract. Not a per-second tariff; do not multiply USD 0.01 by clip length. Failed-request and retry charging are not specified.
Pricing source (opens in a new tab)

Data & enterprise terms

fal documents configurable file lifecycle. The legacy CassetteAI mobile-app privacy policy does not establish hosted API retention, training use or data residency.

Read the service policy (opens in a new tab)

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

fal API

cassetteai/sound-effects-generator
Not verified

No CassetteAI-specific prompt-caching contract.

How it works
Not collected

No CassetteAI-specific prompt-caching contract.

Minimum reusable input
Not collected

No CassetteAI-specific prompt-caching contract.

Lifetime & refresh
Not collected

No CassetteAI-specific prompt-caching contract.

How to enable it
Not collected

No CassetteAI-specific prompt-caching contract.

Usage fields
Not collected

No CassetteAI-specific prompt-caching contract.

Billing rules
Not collected

No CassetteAI-specific prompt-caching contract.

Limits & data boundaries
Not collected

No CassetteAI-specific prompt-caching contract.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Caching reviewed 14 Oct 2026

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Qualified closed-weight inference: the developer operates this named SFX Generator within its proprietary CassetteAI audio platform, with restricted model/source-code rights and commercial hosted/on-device licensing. The terms explicitly name the music model rather than a distinct SFX checkpoint; exact SFX weights and any independent release license remain unverified. Ownership and separate SFX identity are established by the official developer and fal release.

Classification evidence (opens in a new tab)

Lifecycle dates

Music and SFX generators announced on fal27 Mar 2025Source (opens in a new tab)

Sources reviewed