Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Meta / Audiobox evaluation

Joint-CLAP · research embedding model

Joint audio and text embedding model for matching descriptions of speech style and environmental sound.

Embeddings
restrictedClosed weightsEmbeddings
Official documentation (opens in a new tab)Reviewed 13 Oct 2026
6 of 18 specification fields collected
Context windowNot collectedAwaiting source review
Maximum outputNot collectedAwaiting source review
Input billingNot collectedMeta · research reference
Output billingNot collectedNo inferred price

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 output
Not collected
Input modalities
Text, AudioSource for Input modalities (opens in a new tab)
Output modalities
EmbeddingsSource for Output modalities (opens in a new tab)
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected
Streaming
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2023-12-25Source for Release date (opens in a new tab)

Research disclosure date, not general API or weight availability.

Parameters
Not collected
Architecture
Jointly fine-tuned WavLM-base+ audio encoder and RoBERTa-base text encoder, trained with a contrastive objectiveSource for Architecture (opens in a new tab)
Languages
Not collected
Context caching
Not collected

No model-specific inference service or discounted prompt-cache contract is established for these research weights.

Published training data
SD-tag-6K, SD-cap-150 and 2,000 hours of speech with human or automatically generated descriptionsSource for Published training data (opens in a new tab)
Vector normalization
Normalized audio and text embeddings before contrastive similaritySource for Vector normalization (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.

Meta · research reference

Provider source (opens in a new tab)
API model ID
Not collected

Descriptive catalog identity, not a verified selectable API ID.

Access
Paper and examples available. No original public checkpoint or callable inference contract verified.
Regions
Research deployment only; public availability and residency are not established

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

The cited research release does not supply a current production processing/retention agreement. Research access or an expired demonstration is not a production API entitlement.

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.

Meta · research reference

Not verified

No published service cache policy for this research record.

How it works
Not collected

No published service cache policy for this research record.

Minimum reusable input
Not collected

No published service cache policy for this research record.

Lifetime & refresh
Not collected

No published service cache policy for this research record.

How to enable it
Not collected

No published service cache policy for this research record.

Usage fields
Not collected

No published service cache policy for this research record.

Billing rules
Not collected

No published service cache policy for this research record.

Limits & data boundaries
Not collected

No published service cache policy for this research record.

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

What the measurements say

Speech text-to-audio retrieval R@10

Meta research authors · provider-reported
22.01% recall

Audiobox Table 8, authors’ held-out speech-description retrieval task. Not general transcription accuracy.

Evaluation source (opens in a new tab)

Sound text-to-audio retrieval R@10

Meta research authors · provider-reported
67.64% recall

Audiobox Table 8, authors’ sound-description retrieval evaluation, separate from speech.

Evaluation source (opens in a new tab)

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

The Audiobox authors describe their separately trained Joint-CLAP evaluator but provide no original checkpoint download in the reviewed paper/release. Public LAION-CLAP, RoBERTa and WavLM weights are distinct base or comparator models, not this adapted checkpoint.

Classification evidence (opens in a new tab)

Lifecycle dates

Joint-CLAP disclosed in Audiobox paper25 Dec 2023Source (opens in a new tab)

Sources reviewed