/ Meta / Voicebox synthetic-speech classifier · research component Meta / Voicebox
Voicebox synthetic-speech classifier · research component Binary research classifier trained to detect audio generated or edited by Voicebox.
restricted Closed weights Deepfake detection
Media analyzed Audio See specification notes below
Analysis output Classification scores See specification notes below
Input billing Not collected Meta · research reference
Output billing Not collected No inferred price
Specifications Access & pricing Caching Evaluation Sources & lifecycle
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
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
Parameters Not collected
Languages Not collected Context caching Not collected No model-specific inference service or discounted prompt-cache contract is established for these research weights.
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 Not collectedDescriptive 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 INPUT Process & retain The system processes eligible input and retains reusable state.
02 / LATER REQUEST Check available state Reuse depends on matching input, session state and the cache policy.
03 / CACHE HIT Reuse & 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
Resynthesized versus Voicebox detection accuracy · 30% mask Meta research authors · provider-reported 70.4% accuracy Table 8, controlled LibriSpeech dev-clean evaluation with 30% phoneme masking. Does not describe unselected real-world audio.
Evaluation source (opens in a new tab)
Resynthesized versus Voicebox detection accuracy · 90% mask Meta research authors · provider-reported 90.7% accuracy Table 8, controlled LibriSpeech dev-clean evaluation with 90% phoneme masking. Different masking regime from the 30% result.
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 Meta describes this classifier as a research safeguard alongside the unreleased Voicebox system. The reviewed paper and release provide the architecture and evaluation, not a public detector checkpoint or API.
Classification evidence (opens in a new tab) Lifecycle dates Sources reviewed Model notes Nine current/historical Muse identities plus closed-weight research models, trained variants and named pipeline components, including 2026 Text-Audiobox research. API availability is explicitly separated from research references. Regional/private models, historical successors and hosted partner routes remain under review.
Research reference only. No public API selector or model-specific price has been verified. Current Muse tariffs, token limits and processing policies do not apply to this record.
This is not a universal detector for every voice-cloning model. The paper distinguishes easy detection of vocoder artifacts from the harder resynthesized-audio comparison.
Averaging all windows can miss localized edits when most of an utterance is genuine. No production false-positive rate or real-world accuracy guarantee is inferred.
Check against your workload Try representative inputs, measure task success and total cost, then check the exact endpoint’s data policy.
Read our methodology
Also from Meta Muse Wonder Muse Wonder Full Muse Spark 1.3 Muse Spark 1.2 Muse Spark 1.1 Muse Spark · original April release Muse Image Muse Voice Transcribe Muse Video · preview Movie Gen Video Personalized Movie Gen Video Movie Gen Edit