Skip to content

#GitHub

  1. Video modelsReddit StableDiffusion · 创作者与工作流分享

    PAN 2 released: open-source 29B video generation model

    PAN 2 released, an open-source 29B latent diffusion model supporting video generation up to 20 seconds, 720p, 24fps, under Apache 2.0 license. The publisher also released training code and provides integrations with Diffusers, FastVideo, and vLLM. The model enables an LLM to generate dense text states/scripts, timing events and actions, with camera nested under events but decoupled from actions.

  1. ToolsReddit comfyui · 创作与工作流分享

    Mago open-sources an AI video studio running on ComfyUI

    Mago has open-sourced an AI video studio running on ComfyUI, designed for long-form video production such as AI series, short films, and micro-dramas. It supports everything from script to storyboard, shot, and character consistency, can render locally, and supports models including MiniMax H3, Qwen, and FLUX. The current version is 0.19.0.

  1. ToolsReddit comfyui · 创作与工作流分享

    Spool: Generate and compare multiple versions shot by shot using a timeline within ComfyUI

    Spool is a timeline tool that runs inside ComfyUI. In a separate tab, dragging a workflow in turns it into a shot. You can generate multiple candidate versions and compare them frame by frame to select. Unselected versions remain under the shot and can be re-selected at any time. The selected version can be dragged onto the timeline to continue stacking other workflows, without adding a Spool node. Rendering is done on the local GPU or Comfy Cloud, and the final result can be exported to Resolve or Premiere for editing.

  1. ToolsGitHub:AI视频开源项目发现

    Open-source tool anymove released, can convert AI videos to 3D animations

    patch-gg open-sourced anymove, a workflow that converts AI videos into 3D animations, supporting creatures such as dragons, spiders, wolves, and demons, with a cost of about $0.26 per clip. The method uses dual-camera same-frame capture plus joint reader fitting to generate GLB files at 24fps, requiring no manual editing and completing within one minute. The repository also includes Agent Skills and MiniMax H3 LoRA.

  1. Toolspyvideotrans 官方发布

    pyvideotrans v4.17 released

    pyvideotrans v4.17 released, adding the Audio.cpp speech recognition channel, and adding noise reduction and vocal separation models based on audio.cpp. It also fixes the nemotronasr repeated download issue, adds the Opper translation channel, and adds a 'further simplify' option to subtitle translation.

  1. TutorialsReddit StableDiffusion · 创作者与工作流分享

    How to train a MOSS-Image LoRA locally on 8 GB of VRAM: parameters, tips, and workflow

    The author disclosed a set of parameters for training a MOSS-Image character LoRA locally on 8 GB of VRAM: the default setting trains for 800 steps, extended to 1200–1500 steps for large datasets or difficult concepts, with a learning rate of 0.0004, rank/alpha 8/8, and 512² resolution.

  1. Video modelsReddit StableDiffusion · 创作者与工作流分享

    OpenMOSS releases open-source image model MOSS-Image

    OpenMOSS has released the image model MOSS-Image, with SFT and Base versions now available on Hugging Face, along with ComfyUI integration instructions. The model is trained on 3.2B images, uses the Apache 2.0 license, supports JSON and BBOX input but does not require them, and is compatible with both natural language and booru tag prompts.

  1. Video modelsBlack Forest Labs 官方博客

    Black Forest Labs releases FLUX 3 Image Klein open-weight model

    Black Forest Labs releases FLUX 3 Image Klein, a 5.6B-parameter open-weight model that supports text-to-image, image editing, and multi-reference generation with up to 10 reference images, and features native 4K generation, pixel-level editing, and bounding-box control.

  1. Video modelsSuSu_酥酥

    Alibaba Tongyi open-sources Qwen-Image-2.1-Turbo, producing 2K images in 8 steps

    Alibaba's Tongyi open-sources Qwen-Image-2.1-Turbo, an accelerated checkpoint based on Qwen-Image-2.1 that uses the same 7B visual generation architecture; text-to-image and natural-language editing require only 8 denoising steps, and the official release says it can still generate 2K images and supports continuous editing such as adding accessories and changing scenes.

  1. Video modelsGitHub:AI视频开源项目发现

    Kandinsky 6.0 open-sourced: synchronized audio-video generation models, in Lite and Pro tiers

    kandinskylab released the Kandinsky 6.0 Video series of diffusion models, including a 3B-parameter Lite and a 29B-parameter Pro, capable of generating 5-second videos with 44 kHz synchronized audio (including lip-sync), supporting text-to-audio-video and image-to-audio-video, and paired with a super-resolution model that upscales output to Full HD (1920×1080).

Browse all news