Public registry snapshot: 2026-10-03. Enable JavaScript or follow Public model JSON for current registry data.

Audio quality

UTMOS 2022

Speech naturalness prediction using a TorchScript UTMOS 2022 implementation.

SOURCE CATALOG ENTRY

Audio inference is not available here.

Inspect the source’s declared task. This source task has no qualified executable audio adapter supported by this page. No audio input, inference result or inference price is offered by this page.

Metadata exports describe the source; they do not run this model. Follow the source documentation while execution support is unavailable.

Read the pinned source ↗Inspect the live model contract ↗

MODEL PROVENANCE

A source you can trace.

Repository: Blinorot/UTMOS-PyTorch ↗ (pinned source, opens in a new tab)
Source revision: 4f2447e519df3b88567b45583d3500006729502b
Declared licence: mit
Declared artifacts: 1
Available evidence: Source metadata

Licence labels describe the registry declaration. Review the pinned source’s terms for each artifact; commercial eligibility has not been established here.

Artifact hashes are source declarations. Local weight verification, quality, latency and cost measurements are not provided by this catalog.

Upstream implementation ↗

Declaration confidence: primary card declaration. Pinned license declaration ↗ Commercial eligibility: unknown.

Upstream task contract

Source input: Audio decoded and preprocessed according to upstream model documentation.
Source output: Predicted quality features; interpretation depends on the model domain.
Output names: overall
ModelUse example: ModelUse inference recipe unavailable

Upstream documentation ↗

These describe the upstream model. ModelUse inference limits and a working execution example are not published.

What can I do here?

Metadata only in this view

Source task: Audio quality
Audio execution: Audio execution unavailable in this catalog view. A source task label does not establish executable availability.
Executable task here: None supported by this catalog view
Manifest price / cost: No runnable manifest here. Price unavailable. Host and storage cost are unknown.
Audio price / usage: No audio price is published here. No price or free entitlement can be inferred from a source capability.
Manifest input: No manifest input accepted here.
Manifest output: Public source metadata only.

Agent discovery and example

Read this model without an account. The live API is authoritative for currently registered recipes.

Public model JSON ↗API contract ↗