Aller au contenu
PipePrésentéHébergé

LTX 2.3 audio-driven lip-sync video

Generate an audio-driven speaking or singing video from a reference image and a selected segment of uploaded audio, with image-aware prompt expansion, 2× spatial upscaling and refinement. Output defaults to an H.264 MP4 containing the cropped uploaded recording.

Workflow details

FamilleLTX-VideoRésolution1024x1024Estimé VRAMNot yet characterizedÉditeurSweet Tea AutomationMis à jour26 oct.BClasse

Profil d'exécution

Éditeur
Sweet Tea Automation
Statut
Présenté
Version
Checking published version…
Mis à jour
26 oct. 2026

Hover nodes for Pipe controls

Core

Primary Pipe surface

Reference image

LoadImage

Input

Source audio

LoadAudio

Input

Audio segment

AudioCrop

Performance description

CR Text

Output size

easy int

Output proportions

LayerUtility: ImageScaleByAspectRatio V2

Extended

Additional surfaced controls

Negative prompt

CLIPTextEncode

Frame rate

easy int

Style LoRA

LoraLoaderModelOnly

Initial image guidance

LTXVImgToVideoInplace

Refinement image guidance

LTXVImgToVideoInplace

Initial randomization

RandomNoise

Refinement randomization

RandomNoise

Initial sampling

BasicScheduler

Initial prompt guidance

CFGGuider

Initial sampler

KSamplerSelect

Refinement prompt guidance

CFGGuider

Refinement sampler

KSamplerSelect

Refinement noise schedule

ManualSigmas

Distillation

LoraLoaderModelOnly

Reference preprocessing

LTXVPreprocess

Conditioning image size

ResizeImagesByLongerEdge

Reference framing

ResizeImageMaskNode

Prompt expansion

RH_LLMAPI_Pro_Node

Video encoding

SweetTeaPreviewVideoFromImages

Output

What you need

Model files
7.1 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

How this recipe works

Tiled VAE decoding · Lip-sync video with source audio

The saved recipe establishes these conditions

Output

  • Lip-sync video with source audio · image

Exigences

Modèles

5

Packs de nœuds

5

Sources

1 source
Sweet TeaSource