Today's Image & Video Generation: Fastest-Growing Projects — May 16, 2026
Today's the Image & Video Generation space on GitHub, there's a noticeable uptick in projects centered around AI-driven image editing and generation, with several repositories demonstrating robust growth and community engagement. The CookSleep/gpt_image_playground project stands out for its extensive activity and popularity, reflecting growing interest in tools that facilitate interactive experimentation with OpenAI’s gpt-image-2 API.
CookSleep's gpt_image_playground is a repository containing an interface built around the OpenAI gpt-image-2 endpoint, offering users the ability to generate and edit images. Its impressive growth score of 64.39 and 1,285 stars indicate that developers are keen on leveraging this tool for hands-on experience with advanced image generation capabilities.
mrslimslim's gpt-image-canvas provides a local AI canvas using tldraw to facilitate professional-grade editing and creation tasks. With a growth score of 26.58 and 342 stars, the project is gaining traction as users appreciate its potential for localized image manipulation without relying on cloud services.
The ian-handdrawn-ppt repository by helloianneo focuses on generating hand-drawn technical PPT pages in PNG format, specifically designed to fit both 21:9 cover and 16:9 content layouts. Its growth score of 25.40 and 818 stars highlight its appeal among users looking for creative and visually distinctive presentation materials.
EvoLinkAI's awesome-gpt-image-2-API-and-Prompts serves as a comprehensive repository compiling API documentation and prompt examples for the GPT-Image-2 system, aiding developers in understanding how to effectively use these interfaces. With an impressive 14,467 stars and a growth score of 22.07, this resource is becoming indispensable for those working with advanced image generation APIs.
lidge-jun's ima2-gen offers both command-line interface (CLI) and web UI access to OpenAI’s GPT Image 2 service, supporting both paid API key authentication and free OAuth via ChatGPT. The project’s growth score of 18.28 and 215 stars reflect its utility in providing flexible options for text-to-image and image-to-image transformations.
RareSense's Nova3D is an open-source client for a platform enabling editable, part-aware 3D generation from text or reference images. Although star metrics are not provided, the repository’s growth score of 9.50 suggests interest among developers exploring novel methods for 3D content creation and manipulation.
The AnyRecon project by OpenImagingLab is dedicated to arbitrary-view 3D reconstruction using a video diffusion model, aiming at generating realistic 3D scenes from videos or images. With an 8.93 growth score and 335 stars, the repository shows promise in advancing the field of 3D content generation.
Kevin Chin's ImageLens is a Chrome extension designed to reverse-engineer prompts from web images and validate them instantly through live image generation, offering users a powerful way to explore and refine their image generation workflows. Its growth score of 5.44 and 66 stars indicate steady interest among developers looking for practical tools in this space.
While ailuntx's Thinking-with-Visual-Primitives is an archived snapshot with limited active development (growth score: 5.34, 39 stars), it still garners attention from researchers interested in the theoretical underpinnings of visual primitive-based AI generation techniques.
Lastly, jieyefriic's nbcraft provides a multi-backend CLI for image and video generation, supporting various services like Gemini, DashScope, Volcengine Ark, and OpenAI gpt-image-2. With a growth score of 5.31 and 149 stars, the project is attracting users seeking versatile and efficient tools for their creative projects.
These repositories collectively demonstrate the dynamic nature of AI-driven image and video generation, with developers actively contributing to and utilizing these tools for both practical applications and theoretical exploration in the field.
CookSleep's gpt_image_playground is a repository containing an interface built around the OpenAI gpt-image-2 endpoint, offering users the ability to generate and edit images. Its impressive growth score of 64.39 and 1,285 stars indicate that developers are keen on leveraging this tool for hands-on experience with advanced image generation capabilities.
mrslimslim's gpt-image-canvas provides a local AI canvas using tldraw to facilitate professional-grade editing and creation tasks. With a growth score of 26.58 and 342 stars, the project is gaining traction as users appreciate its potential for localized image manipulation without relying on cloud services.
The ian-handdrawn-ppt repository by helloianneo focuses on generating hand-drawn technical PPT pages in PNG format, specifically designed to fit both 21:9 cover and 16:9 content layouts. Its growth score of 25.40 and 818 stars highlight its appeal among users looking for creative and visually distinctive presentation materials.
EvoLinkAI's awesome-gpt-image-2-API-and-Prompts serves as a comprehensive repository compiling API documentation and prompt examples for the GPT-Image-2 system, aiding developers in understanding how to effectively use these interfaces. With an impressive 14,467 stars and a growth score of 22.07, this resource is becoming indispensable for those working with advanced image generation APIs.
lidge-jun's ima2-gen offers both command-line interface (CLI) and web UI access to OpenAI’s GPT Image 2 service, supporting both paid API key authentication and free OAuth via ChatGPT. The project’s growth score of 18.28 and 215 stars reflect its utility in providing flexible options for text-to-image and image-to-image transformations.
RareSense's Nova3D is an open-source client for a platform enabling editable, part-aware 3D generation from text or reference images. Although star metrics are not provided, the repository’s growth score of 9.50 suggests interest among developers exploring novel methods for 3D content creation and manipulation.
The AnyRecon project by OpenImagingLab is dedicated to arbitrary-view 3D reconstruction using a video diffusion model, aiming at generating realistic 3D scenes from videos or images. With an 8.93 growth score and 335 stars, the repository shows promise in advancing the field of 3D content generation.
Kevin Chin's ImageLens is a Chrome extension designed to reverse-engineer prompts from web images and validate them instantly through live image generation, offering users a powerful way to explore and refine their image generation workflows. Its growth score of 5.44 and 66 stars indicate steady interest among developers looking for practical tools in this space.
While ailuntx's Thinking-with-Visual-Primitives is an archived snapshot with limited active development (growth score: 5.34, 39 stars), it still garners attention from researchers interested in the theoretical underpinnings of visual primitive-based AI generation techniques.
Lastly, jieyefriic's nbcraft provides a multi-backend CLI for image and video generation, supporting various services like Gemini, DashScope, Volcengine Ark, and OpenAI gpt-image-2. With a growth score of 5.31 and 149 stars, the project is attracting users seeking versatile and efficient tools for their creative projects.
These repositories collectively demonstrate the dynamic nature of AI-driven image and video generation, with developers actively contributing to and utilizing these tools for both practical applications and theoretical exploration in the field.