Today's Image & Video Generation: Fastest-Growing Projects — August 27, 2026
Today's the Image & Video Generation space on GitHub, we see a continued surge in projects that leverage advanced AI techniques to create and manipulate images and videos with high fidelity and artistic flair. Among these, the project `ant-research/4DAnyone` stands out as one of the most rapidly growing initiatives, focusing on generating realistic 3D models from casual monocular video inputs.
The `ant-research/4DAnyone` repository has seen significant traction this week, with a growth score of 31.85 and accumulating over 810 stars. This project aims to create detailed 4D representations of individuals or objects by analyzing casual monocular videos, showcasing its potential for applications in virtual reality and gaming.
The `Cripacx/mediagen` tool is designed to streamline the process of generating AI-powered images and videos through a command-line interface (CLI), integrating multiple AI services such as Gemini, OpenAI, and Kie AI. With 29 points on our growth radar and nearly 56 stars, this project highlights its versatility in handling various AI tasks under one unified platform.
`shanliuling/dsh-image-gen`, with a growth score of 24.40 and 221 stars, offers an innovative way to generate images directly within DeepSeek Harness chats. This tool leverages local access to neural networks for image generation, making it particularly appealing for users who prefer on-device processing.
`six-nut/PocketMen-with-you` is a project that allows users to create high-fidelity companions from two or more photos using open-weight neural editing techniques without requiring an OpenAI API key. With 14.40 points in growth and over 150 stars, this tool demonstrates the potential for local AI-based image processing solutions.
The `dacnay816y62-hub/photo-revival` repository is focused on transforming everyday photos into poetic, hand-drawn illustrations through Codex skills. This project has garnered considerable attention with a growth score of 13.19 and 497 stars, reflecting its unique approach to image transformation.
`DandreYang/local-image-gen`, boasting a growth score of 13.11 and 53 stars, provides users the ability to generate images from local Grok, Antigravity, Cursor, or Codex logins, offering an alternative to cloud-based API services with official API keys as backup options.
`csslc/PixRestore`, which has a growth score of 9.70 and 78 stars, is dedicated to image restoration through the Pixel Diffusion Transformer technique. This project highlights its technical depth in handling complex image restoration tasks.
The `yokel1121/muyang-illustration-skills` repository offers a diverse set of illustration styles for Codex, including prompt recipes and visual references, with 9 skills available to users. With a growth score of 9.25 and 191 stars, this project showcases the growing demand for creative AI-driven image generation tools.
`Sateezg/codex-bridge`, featuring a growth score of 8.42 and 249 stars, provides GPT-5 subagents for Claude Code through the Codex CLI login without requiring an OpenAI API key, making it accessible to users with existing Codex environments.
Lastly, `moskoo/xxg-portrait-rebuild-light` is a portrait reconstruction tool that enhances lighting and texture in existing photos. With 7.41 points on our growth radar and 120 stars, this project demonstrates the importance of detailed image editing tools for professional photography and design needs.
The `ant-research/4DAnyone` repository has seen significant traction this week, with a growth score of 31.85 and accumulating over 810 stars. This project aims to create detailed 4D representations of individuals or objects by analyzing casual monocular videos, showcasing its potential for applications in virtual reality and gaming.
The `Cripacx/mediagen` tool is designed to streamline the process of generating AI-powered images and videos through a command-line interface (CLI), integrating multiple AI services such as Gemini, OpenAI, and Kie AI. With 29 points on our growth radar and nearly 56 stars, this project highlights its versatility in handling various AI tasks under one unified platform.
`shanliuling/dsh-image-gen`, with a growth score of 24.40 and 221 stars, offers an innovative way to generate images directly within DeepSeek Harness chats. This tool leverages local access to neural networks for image generation, making it particularly appealing for users who prefer on-device processing.
`six-nut/PocketMen-with-you` is a project that allows users to create high-fidelity companions from two or more photos using open-weight neural editing techniques without requiring an OpenAI API key. With 14.40 points in growth and over 150 stars, this tool demonstrates the potential for local AI-based image processing solutions.
The `dacnay816y62-hub/photo-revival` repository is focused on transforming everyday photos into poetic, hand-drawn illustrations through Codex skills. This project has garnered considerable attention with a growth score of 13.19 and 497 stars, reflecting its unique approach to image transformation.
`DandreYang/local-image-gen`, boasting a growth score of 13.11 and 53 stars, provides users the ability to generate images from local Grok, Antigravity, Cursor, or Codex logins, offering an alternative to cloud-based API services with official API keys as backup options.
`csslc/PixRestore`, which has a growth score of 9.70 and 78 stars, is dedicated to image restoration through the Pixel Diffusion Transformer technique. This project highlights its technical depth in handling complex image restoration tasks.
The `yokel1121/muyang-illustration-skills` repository offers a diverse set of illustration styles for Codex, including prompt recipes and visual references, with 9 skills available to users. With a growth score of 9.25 and 191 stars, this project showcases the growing demand for creative AI-driven image generation tools.
`Sateezg/codex-bridge`, featuring a growth score of 8.42 and 249 stars, provides GPT-5 subagents for Claude Code through the Codex CLI login without requiring an OpenAI API key, making it accessible to users with existing Codex environments.
Lastly, `moskoo/xxg-portrait-rebuild-light` is a portrait reconstruction tool that enhances lighting and texture in existing photos. With 7.41 points on our growth radar and 120 stars, this project demonstrates the importance of detailed image editing tools for professional photography and design needs.