输入一个主题，3 分钟吐出一支带配音的短视频
不会剪视频的人有福了：给它一个主题，它自己写文案、配图、配音、加音乐、合成出片。
Pixelle-Video，阿里国际团队开源的“全自动短视频引擎”，上线不到一年冲到近 3 万星，还有 Windows 整合包，下载解压就能用。
ATH-MaaS/Pixelle-Video
3 likes · 1 repost · 329 views
PixText — модель для правки текста на готовых
картинках Небольшая модель (0,95 млрд параметров), которая исправляет криво сгенерированный текст прямо...
ATH-MaaS/PixText
huggingface.co
registry.comfy.org
Pixelle MCP - Multimodal AIGC Solution | Open-Source & Easy to Use
Pixelle MCP is an open-source multimodal AIGC solution based on ComfyUI + MCP + LLM. It supports both local ComfyUI and cloud RunningHub modes, and offers zero-code development. Explore the project with over one thousand stars and one hundred fifty-three forks. Visit the GitHub page for more details. #AIGC #ComfyUI #OpenSource
🔗 ATH-MaaS/Pixelle-MCP
53 views
PixText: a small, fast model for fixing text in generated images
I built PixText for a very specific frustration: an image looks great, most of the text is right, but a few mangled words or Chinese characters keep me from being happy with it.
VAE compression can lose fine details in small text regions, so I trained the model directly in pixel space. I also wanted the refiner to be small and fast: it's under 1B parameters (about 0.95B), needs no text encoder or VAE, and uses MeanFlow to refine text in just 4 sampling steps.
If you'd like to try it:
- ComfyUI node on the Registry
- Ready-to-use ComfyUI workflows
- Code and installation
- Model weights
I'd love to hear what you think. Feedback and examples are very welcome! 🤗
295 points · 33 comments
请帮我搭建一套开源、免费、本地运行的
「AI 儿童英语绘本动画生成系统」。
优先研究并复用以下 GitHub 项目：
1、ComfyUI（AI 绘图与视频工作流）
Comfy-Org/ComfyUI
2、Pixelle-Video（自动化短视频制作）
ATH-MaaS/Pixelle-Video
3、Wan2.2（AI 动画生成）
Wan-Video/Wan2.2
4、Piper TTS（本地英语配音）
OHF-Voice/piper1-gpl
目标：输入一句故事主题，
自动生成适合3～6岁儿童的英语动画。
具体要求：
1、AI 自动编写英语故事、角色对话和分镜。
2、生成统一画风的3D卡通动物角色。
3、让角色走路、说话、互动，保持形象一致。
4、自动生成英语配音及中英双语字幕。
5、自动添加背景音乐，合成60秒动画。
6、输出1080×1920竖屏MP4，适配抖音。
7、支持批量生成，并提供中文操作界面。
执行方式：
先检查电脑配置、项目功能及开源许可证，
优先复用现成代码，不重复造轮子。
禁止调用付费API，优先使用本地免费模型
模型和素材必须符合实际使用及商用授权要求
先制作一条10秒测试视频，
成功后再生成60秒完整版。
如果硬件不足，自动推荐轻量级替代方案
现在开始检查环境、制定方案并实施
6 likes · 319 views
I Typed 1 Sentence and AI Made This Whole Short
I typed one sentence. This free, open-source AI made the whole Short: script, visuals, voiceover and music. 🤯
🔗 Pixelle-Video on GitHub: ATH-MaaS/Pixelle-Video
How it works:
1. Type any topic
2. An AI model writes the script and plans a shot per sentence
3. It generates an image or video clip for each scene
4. Adds a voiceover + background music
5. Exports a vertical video, ready to post
Open source (Apache-2.0). The software is free; some AI image/video/voice models you plug in may have their own costs.
Would you post a Short made like this? 👇
#shorts #ai #aitools #facelesschannel #youtubeautomation #opensource #github #aivideo
1 view
做短视频最耗时间的环节，现在 AI 想一口气全包了。
Pixelle-Video 开源项目，从一句想法出发，自动生成脚本、分镜、画面、配音、字幕，直到视频合成，把整条创作流程串起来。
支持本地部署和二次开发，做 AI 视频、批量内容的可以重点关注。
🔗 ATH-MaaS/Pixelle-Video t.co
3 likes · 218 views
Mentions
输入一个主题，3 分钟吐出一支带配音的短视频 不会剪视频的人有福了：给它一个主题，它自己写文案、配图、配音、加音乐、合成出片。 Pixelle-Video，阿里国际团队开源的“全自动短视频引擎”，上线不到一年冲到近 3 万星，还有 Windows 整合包，下载解压就能用。 ATH-MaaS/Pixelle-Video
PixText — модель для правки текста на готовых картинках Небольшая модель (0,95 млрд параметров), которая исправляет криво сгенерированный текст прямо... ATH-MaaS/PixText huggingface.co registry.comfy.org
Pixelle MCP - Multimodal AIGC Solution | Open-Source & Easy to Use Pixelle MCP is an open-source multimodal AIGC solution based on ComfyUI + MCP + LLM. It supports both local ComfyUI and cloud RunningHub modes, and offers zero-code development. Explore the project with over one thousand stars and one hundred fifty-three forks. Visit the GitHub page for more details. #AIGC #ComfyUI #OpenSource 🔗 ATH-MaaS/Pixelle-MCP
PixText: a small, fast model for fixing text in generated images I built PixText for a very specific frustration: an image looks great, most of the text is right, but a few mangled words or Chinese characters keep me from being happy with it. VAE compression can lose fine details in small text regions, so I trained the model directly in pixel space. I also wanted the refiner to be small and fast: it's under 1B parameters (about 0.95B), needs no text encoder or VAE, and uses MeanFlow to refine text in just 4 sampling steps. If you'd like to try it: - ComfyUI node on the Registry - Ready-to-use ComfyUI workflows - Code and installation - Model weights I'd love to hear what you think. Feedback and examples are very welcome! 🤗
请帮我搭建一套开源、免费、本地运行的 「AI 儿童英语绘本动画生成系统」。 优先研究并复用以下 GitHub 项目： 1、ComfyUI（AI 绘图与视频工作流） Comfy-Org/ComfyUI 2、Pixelle-Video（自动化短视频制作） ATH-MaaS/Pixelle-Video 3、Wan2.2（AI 动画生成） Wan-Video/Wan2.2 4、Piper TTS（本地英语配音） OHF-Voice/piper1-gpl 目标：输入一句故事主题， 自动生成适合3～6岁儿童的英语动画。 具体要求： 1、AI 自动编写英语故事、角色对话和分镜。 2、生成统一画风的3D卡通动物角色。 3、让角色走路、说话、互动，保持形象一致。 4、自动生成英语配音及中英双语字幕。 5、自动添加背景音乐，合成60秒动画。 6、输出1080×1920竖屏MP4，适配抖音。 7、支持批量生成，并提供中文操作界面。 执行方式： 先检查电脑配置、项目功能及开源许可证， 优先复用现成代码，不重复造轮子。 禁止调用付费API，优先使用本地免费模型 模型和素材必须符合实际使用及商用授权要求 先制作一条10秒测试视频， 成功后再生成60秒完整版。 如果硬件不足，自动推荐轻量级替代方案 现在开始检查环境、制定方案并实施
I Typed 1 Sentence and AI Made This Whole Short I typed one sentence. This free, open-source AI made the whole Short: script, visuals, voiceover and music. 🤯 🔗 Pixelle-Video on GitHub: ATH-MaaS/Pixelle-Video How it works: 1. Type any topic 2. An AI model writes the script and plans a shot per sentence 3. It generates an image or video clip for each scene 4. Adds a voiceover + background music 5. Exports a vertical video, ready to post Open source (Apache-2.0). The software is free; some AI image/video/voice models you plug in may have their own costs. Would you post a Short made like this? 👇 #shorts #ai #aitools #facelesschannel #youtubeautomation #opensource #github #aivideo
做短视频最耗时间的环节，现在 AI 想一口气全包了。 Pixelle-Video 开源项目，从一句想法出发，自动生成脚本、分镜、画面、配音、字幕，直到视频合成，把整条创作流程串起来。 支持本地部署和二次开发，做 AI 视频、批量内容的可以重点关注。 🔗 ATH-MaaS/Pixelle-Video t.co