Today's Image & Video Generation: Fastest-Growing Projects — August 16, 2026
Today's the Image & Video Generation space on GitHub, we see a mix of innovative projects that range from artistic illustration skills and context-bounded image generation to advanced video editing capabilities. Notable repositories include tools for turning everyday photos into hand-drawn illustrations and plugins that enable deep understanding of images locally and privately.
yokel1121/muyang-illustration-skills offers 25 fixed illustration styles and prompt recipes through Codex, enhancing the creative process for artists. With a growth score of 28.30 and 178 stars, its detailed resources and broad appeal to illustrators are driving interest.
Sateezg/codex-bridge provides image generation capabilities and GPT-5 subagents through the Codex CLI without requiring an OpenAI API key. This accessibility and utility across various projects have contributed to its growth score of 23.29 and 272 stars.
shixinnt/codex-image-context-runtime is a plugin for local context-bounded image generation and inspection, offering significant privacy benefits by avoiding cloud dependencies. Its recent activity with 18 commits in the last month and a growth score of 21.40 reflect its growing importance among developers seeking on-device solutions.
dacnay816y62-hub/photo-revival transforms everyday photos into poetic, hand-drawn illustrations using Codex skills. With 418 stars and a growth score of 19.53, the artistic appeal and potential for creative expression are evident in its rising popularity.
oil-oil/dsh-vision aims to provide near-native image understanding capabilities for DeepSeek Harness, enhancing how images are processed and analyzed in various applications. The project's modest but steady growth with a score of 17.25 and 39 stars suggests ongoing development that could attract more users as it matures.
moskoo/xxg-portrait-rebuild-light offers an image-edit skill for reconstructing portrait lighting and skin texture, using a director-led workflow to enhance photo realism. With 44 stars and a growth score of 11.60, its specialized focus on portrait photography is gaining traction among photographers and enthusiasts.
Tencent-Hunyuan/Hunyuan3D-Buffalo1.0 presents a unified multimodal model for scalable 3D generation, understanding, and editing, showcasing the potential of advanced 3D technology in various applications. Its growing interest, with 7 commits this month and a growth score of 8.38, highlights its relevance to developers working on complex visual projects.
mtsee/unicut is an AI-driven video editing tool that simplifies the process for content creators looking to produce professional-quality videos efficiently. With 183 stars and a growth score of 7.96, its active development (24 commits in the last month) signals strong community engagement and continued improvement.
itsmeadarsh2008/opencode-senses integrates vision capabilities into OpenCode, allowing users to analyze images with deep understanding locally and privately. The high number of recent commits (51 this month), coupled with a growth score of 6.59 and 60 stars, suggests strong community involvement and rapid development.
BeatAPI/awesome-minimax-h3-prompts curates MiniMax H3 video generation prompts across various styles, including cinematic, ads, anime, UGC, product videos, and more. Its utility for content creators looking to generate diverse types of videos is reflected in its 104 stars and a growth score of 6.53.
These repositories showcase the dynamic landscape of image and video generation tools, with projects ranging from artistic creativity to technical innovation and practical applications like video editing and privacy-focused image analysis.
yokel1121/muyang-illustration-skills offers 25 fixed illustration styles and prompt recipes through Codex, enhancing the creative process for artists. With a growth score of 28.30 and 178 stars, its detailed resources and broad appeal to illustrators are driving interest.
Sateezg/codex-bridge provides image generation capabilities and GPT-5 subagents through the Codex CLI without requiring an OpenAI API key. This accessibility and utility across various projects have contributed to its growth score of 23.29 and 272 stars.
shixinnt/codex-image-context-runtime is a plugin for local context-bounded image generation and inspection, offering significant privacy benefits by avoiding cloud dependencies. Its recent activity with 18 commits in the last month and a growth score of 21.40 reflect its growing importance among developers seeking on-device solutions.
dacnay816y62-hub/photo-revival transforms everyday photos into poetic, hand-drawn illustrations using Codex skills. With 418 stars and a growth score of 19.53, the artistic appeal and potential for creative expression are evident in its rising popularity.
oil-oil/dsh-vision aims to provide near-native image understanding capabilities for DeepSeek Harness, enhancing how images are processed and analyzed in various applications. The project's modest but steady growth with a score of 17.25 and 39 stars suggests ongoing development that could attract more users as it matures.
moskoo/xxg-portrait-rebuild-light offers an image-edit skill for reconstructing portrait lighting and skin texture, using a director-led workflow to enhance photo realism. With 44 stars and a growth score of 11.60, its specialized focus on portrait photography is gaining traction among photographers and enthusiasts.
Tencent-Hunyuan/Hunyuan3D-Buffalo1.0 presents a unified multimodal model for scalable 3D generation, understanding, and editing, showcasing the potential of advanced 3D technology in various applications. Its growing interest, with 7 commits this month and a growth score of 8.38, highlights its relevance to developers working on complex visual projects.
mtsee/unicut is an AI-driven video editing tool that simplifies the process for content creators looking to produce professional-quality videos efficiently. With 183 stars and a growth score of 7.96, its active development (24 commits in the last month) signals strong community engagement and continued improvement.
itsmeadarsh2008/opencode-senses integrates vision capabilities into OpenCode, allowing users to analyze images with deep understanding locally and privately. The high number of recent commits (51 this month), coupled with a growth score of 6.59 and 60 stars, suggests strong community involvement and rapid development.
BeatAPI/awesome-minimax-h3-prompts curates MiniMax H3 video generation prompts across various styles, including cinematic, ads, anime, UGC, product videos, and more. Its utility for content creators looking to generate diverse types of videos is reflected in its 104 stars and a growth score of 6.53.
These repositories showcase the dynamic landscape of image and video generation tools, with projects ranging from artistic creativity to technical innovation and practical applications like video editing and privacy-focused image analysis.