Lip Forcing

6.4easy start · #15 of 27
in AI Video Lip Sync Tools

Free to practise onnot on record
Free trialnot on record
Well documentedyes
Runs where you worknot on record

Runs on Linux, self-hosted.

Visit site Alternatives

Lip Forcing is ranked #15 of 27 in AI video lip sync tools on The Geeks Club. It runs on Linux, Self-hosted.

Compared on AI video lip sync tools

Free plan: Yes

Best Lip Forcing alternatives

See all 20

CHAMELAION LipSync API 0 of 4 in the starter kit Lesson 17.7 LipDub 0 of 4 in the starter kit Lesson 27.6 Sync Labs 0 of 4 in the starter kit Lesson 37.6 VEED 0 of 4 in the starter kit Lesson 47.3 Morph Studio 0 of 4 in the starter kit Lesson 57.2 lipsyn.cc 0 of 4 in the starter kit Lesson 67.1

The record

Purpose: Lip Forcing is an autoregressive diffusion method for video-to-video lip synchronization that generates lip-synced video from a reference video and audio.
Generation: Its causal student models generate each chunk in two denoising steps without classifier-free guidance at inference.
Streaming: Streaming inference processes chunks on the fly, with sub-millisecond time-to-first-frame and GPU memory reported as constant with clip length under the stated setup.
Performance: The project reports that the 1.3B student reaches 31 FPS and the 14B student runs 39.8 times faster than its teacher at comparable reference fidelity.
Inference setup: The documented inference command takes a reference video and speech audio and writes an output video.
Automatic processing: The inference pipeline automatically detects and aligns faces to 512×512 crops and pastes the generated result back into the video.
Weights: The released 14B student checkpoint is a merged, self-contained file, while the 1.3B student weights are listed as coming soon.
External components: Inference also requires external components including the Wan VAE, wav2vec audio encoder, UMT5-XXL text encoder or precomputed embeddings, TAEW decoder, and mouth mask.
Hardware: The repository reports testing on Ubuntu 24.04 with an NVIDIA H200 and says inference runs on a single GPU.
Resource limit: For the 14B model, the repository reports about 37 GB peak VRAM with precomputed text embeddings and about 50 GB with runtime text encoding.
Training: Training is documented as two stages: Diffusion-Forcing initialization followed by Self-Forcing DMD with a SyncNet reward.
License: The code repository states that Lip Forcing is released under the Apache License 2.0.
Hosting: The checkpoint page states that the model is not deployed by any Hugging Face Inference Provider.
Project affiliation: The project lists KAIST AI and AIPARK affiliations for its authors.

Where it ranks on The Geeks Club

Best AI Video Lip Sync Tools in 2026#15 of 27

Is Lip Forcing yours?

Claim it for free: prove the domain, then correct facts, plans and screenshots. An editor reviews every change.

Claim it free Promote Lip Forcing

Sources

github.com/cvlab-kaist/LipForcing· checked 4 Oct 2026
cvlab-kaist.github.io/LipForcing/· checked 4 Oct 2026
huggingface.co/JinhyukJang/lipforcing· checked 4 Oct 2026