Taming Stable Diffusion for Lip Sync!

🤘 TT-NN operator library, and TT-Metalium low level kernel programming model.

VEFX-Bench: A Holistic Benchmark for Generic Video Editing and Visual Effects

An open source community implementation of the model from the paper: "Movie Gen: A Cast of Media Foundation Models". Join our community to help implement this model!

[ICCV 2025] FiVE-Bench: A Fine-grained Video Editing Benchmark for Evaluating Emerging Diffusion and Rectified Flow Models

Easy recipes to speed up latency of Flux, QwenImage, and LTX-2 with NVFP4 and MXFP8 on Blackwell.

Once the Adobe Sora API can be use, this repository will be updated soon. My other projects: https://videosora.app https://makeimage.ai

使用AI大模型，一键生成高清故事短视频。Generate high-definition story short videos with one click using AI large models.

Create a Waveform Video (usable on Youtube, Tiktok, etc.) from a WAV or MP3 file. Two output options: ultrafast generation (static background with optional title) and standard generation (dynamic background).

🛠️ Synchronize X-UI subscriptions and tunnel inbounds with SQLite, managing traffic, expiry, and user limits automatically every 30 seconds.

AI Video Agent that helps streamline workflows for next-gen video interactions and workflows

Copilot agent for Video Generation

Sync Havoc operator actions to Ghostwriter via GraphQL API for automated, real-time activity logging and streamlined reporting.

Sync Obsidian vaults to GitHub with optional end-to-end encryption, conflict resolution, and restore history.