Today's Image & Video Generation: Fastest-Growing Projects — April 21, 2026
This week, the Image & Video Generation space on GitHub saw a surge in activity around multimodal foundation models and text-to-image generation tools. Several repositories gained significant traction, showcasing the growing interest in AI-powered image and video creation. Notably, many of these projects have achieved impressive growth scores, indicating their potential to shape the future of visual content generation.
jd-opensource/JoyAI-Image (Growth Score: 60.86, Stars: 1,932) is a unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing. Its impressive growth score can be attributed to its comprehensive approach to image processing and generation, making it an attractive solution for developers seeking a versatile tool.
baidu/ERNIE-Image (Growth Score: 34.93, Stars: 303) is an open text-to-image generation model developed by the ERNIE-Image team at Baidu, built on a single-stream Diffusion Transformer (DiT). Its growth can be attributed to its state-of-the-art performance among open-weight text-to-image models, making it a popular choice for researchers and developers.
ashim-hq/ashim (Growth Score: 32.98, Stars: 842) is a Docker container that provides over 30 tools for image processing, including resizing, compression, background removal, upscaling, OCR, and more. Its growth can be attributed to its convenience and flexibility, offering users a self-contained solution for various image processing tasks without relying on cloud services.
kangarooking/design-image-studio (Growth Score: 29.67, Stars: 73) is a tool that converts vague visual requirements into high-quality design images. Although its description is brief, its growth score suggests increasing interest in this specific application of image generation technology.
OpenMOSS/MOSS-VL (Growth Score: 17.69, Stars: 226) is the core multimodal model series within the OpenMOSS ecosystem, dedicated to visual understanding. Its growth can be attributed to its focus on visual understanding, a critical aspect of image and video generation.
ShandaAI/AlayaRenderer (Growth Score: 16.42, Stars: 567) is an AI-native Renderer for Games and Virtual Worlds, designed to generate immersive environments. Its growth can be attributed to the increasing demand for high-quality rendering engines in the gaming and virtual reality industries.
F-R-L/forge-film (Growth Score: 13.50, Stars: 501) is a multi-model DAG-driven parallel AI film generation engine that generates film scenes simultaneously instead of one by one. Its growth can be attributed to its innovative approach to film generation, offering significant speedup and efficiency improvements.
gulucaptain/Camera-Transformer-1 (Growth Score: 10.95, Stars: 100) understands user intent and generates videos with precise, spatially-aware camera control. Its growth can be attributed to the increasing interest in AI-powered video creation tools that offer more control over camera movements and scene composition.
NVlabs/PixelDiT (Growth Score: 3.80, Stars: 140) is a Pixel Diffusion Transformers model for image generation, although its brief description limits further analysis. Nevertheless, its presence on this list indicates ongoing research in diffusion-based image generation techniques.
jd-opensource/JoyAI-Image (Growth Score: 60.86, Stars: 1,932) is a unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing. Its impressive growth score can be attributed to its comprehensive approach to image processing and generation, making it an attractive solution for developers seeking a versatile tool.
baidu/ERNIE-Image (Growth Score: 34.93, Stars: 303) is an open text-to-image generation model developed by the ERNIE-Image team at Baidu, built on a single-stream Diffusion Transformer (DiT). Its growth can be attributed to its state-of-the-art performance among open-weight text-to-image models, making it a popular choice for researchers and developers.
ashim-hq/ashim (Growth Score: 32.98, Stars: 842) is a Docker container that provides over 30 tools for image processing, including resizing, compression, background removal, upscaling, OCR, and more. Its growth can be attributed to its convenience and flexibility, offering users a self-contained solution for various image processing tasks without relying on cloud services.
kangarooking/design-image-studio (Growth Score: 29.67, Stars: 73) is a tool that converts vague visual requirements into high-quality design images. Although its description is brief, its growth score suggests increasing interest in this specific application of image generation technology.
OpenMOSS/MOSS-VL (Growth Score: 17.69, Stars: 226) is the core multimodal model series within the OpenMOSS ecosystem, dedicated to visual understanding. Its growth can be attributed to its focus on visual understanding, a critical aspect of image and video generation.
ShandaAI/AlayaRenderer (Growth Score: 16.42, Stars: 567) is an AI-native Renderer for Games and Virtual Worlds, designed to generate immersive environments. Its growth can be attributed to the increasing demand for high-quality rendering engines in the gaming and virtual reality industries.
F-R-L/forge-film (Growth Score: 13.50, Stars: 501) is a multi-model DAG-driven parallel AI film generation engine that generates film scenes simultaneously instead of one by one. Its growth can be attributed to its innovative approach to film generation, offering significant speedup and efficiency improvements.
gulucaptain/Camera-Transformer-1 (Growth Score: 10.95, Stars: 100) understands user intent and generates videos with precise, spatially-aware camera control. Its growth can be attributed to the increasing interest in AI-powered video creation tools that offer more control over camera movements and scene composition.
NVlabs/PixelDiT (Growth Score: 3.80, Stars: 140) is a Pixel Diffusion Transformers model for image generation, although its brief description limits further analysis. Nevertheless, its presence on this list indicates ongoing research in diffusion-based image generation techniques.