Skip to content

FreeVideo v0.3.0

Choose a tag to compare

View all tags
@github-actions github-actions released this 06 Oct 23:31
· 37 commits to main since this release
v0.3.0
355c7d8
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
FreeVideo

Download / 下载
Windows 10/11 · NVIDIA GPU FreeVideo.exe
macOS 14+ · Apple silicon FreeVideo-Mac-arm64.dmg

Windows: run FreeVideo.exe, choose an existing ComfyUI folder or install a new one, then click Install & launch. Offline packages are on Quark.

macOS: open the DMG and drag FreeVideo.app into Applications. If macOS asks you to confirm the first launch, see the first-open steps.

Windows： 运行 FreeVideo.exe，选择已有的 ComfyUI 或安装新的 ComfyUI，点击 安装并启动。离线包见夸克网盘。

macOS： 打开 DMG，将 FreeVideo.app 拖入「应用程序」。首次打开如需确认，请参考首次打开步骤。

What's new

A faster int8 model for GeForce and RTX 30-series cards with a one-click upgrade, and optional prompt enhancement.

  • GeForce RTX 40/50-series and RTX 30-series cards use a faster int8 model, closer to the original quality.
  • Optional prompt enhancement in the creation panel: a local model rewrites your prompt for MiniMax H3, and you choose the original or the enhanced version. It is off by default and asks before downloading its model (8.3 GiB).
  • Faster generation on 8 GB cards.
  • Failed installations and updates can be retried in place, and download caches can be cleaned from Settings.

更新内容

GeForce 和 RTX 30 系显卡改用更快的 int8 模型，老用户一键升级；新增可选的提示词增强。

  • GeForce RTX 40/50 系和 RTX 30 系显卡换用更快的 int8 模型，画质也更接近原版。
  • 创作面板新增可选的提示词增强：用本地模型把提示词改写成适合 MiniMax H3 的写法，原文和改写版可任选其一生成。默认关闭，首次使用会先询问再下载模型（8.3 GiB）。
  • 8 GB 显卡生成更快。
  • 安装或更新失败可直接重试，设置中可以清理下载缓存。
sha256 
8f0bd5cb936a3b6592b73941aebad607e5cd4b433c69592697ec43ebf9662e75  FreeVideo.exe
4b3d8bd6195bc8bc4f9ca8d2add0add20c949af36ed7f1686929cc8fe985e1f2  FreeVideo-Mac-arm64.dmg

Discord · QQ Group / QQ 群 · WeChat Group / 微信群

Assets 7
Loading