screencastgen

Lipsync Registry

Registry for lip-sync video generation providers.

Source: screencastgen/providers/lipsync/__init__.py, screencastgen/providers/lipsync/base.py

Overview

Manages lip-sync providers that take audio + reference video and produce a lip-synced face video.

Functions

Function	Description
`get_lipsync_provider_names() -> List[str]`	Returns `["auto", "latentsync"]`
`get_default_lipsync_provider() -> str`	Returns `"auto"`
`get_auto_lipsync_provider() -> str`	Returns the first registered concrete provider (`"latentsync"`)
`get_lipsync_provider_spec(name) -> LipsyncProviderSpec`	Get provider spec
`run_lipsync_provider(provider, video_path, audio_path, output_path, device, **kwargs)`	Dispatch to named provider

The "auto" value resolves to the first registered concrete provider. It currently selects LatentSync and reports a configuration error if that runtime is unavailable; it does not fall back to an unregistered implementation.

Dataclass: `LipsyncProviderSpec`

Defined in base.py. Metadata for a lip-sync provider.

Registered Providers

Name	Module	Description
`auto`	—	Select the first registered provider (currently LatentSync)
`latentsync`	LatentSync Provider	High-quality lip-sync

Default: auto

Dependencies

Lipsync Registry
├── LatentSync Provider  (lazy import)
├── LipsyncProviderSpec      (base.py)
└──▶ consumed by Lipsync Facade
     └──▶ consumed by Lipsync Pipeline