Today's Image & Video Generation: Fastest-Growing Projects — April 20, 2026
Today's Image & Video Generation, we're seeing a surge in interest around multimodal foundation models and text-to-image generation tools. Repositories like JoyAI-Image and ERNIE-Image are leading the charge, with significant growth in stars and commits over the past 30 days. Meanwhile, other tools like ashim and AlayaRenderer are gaining traction for their unique approaches to image processing and rendering.
JoyAI-Image, a unified multimodal foundation model for image understanding and generation, has seen an impressive growth score of 63.83 and now boasts 1,929 stars. Its popularity can be attributed to its versatility in handling various tasks such as text-to-image generation and instruction-guided image editing. With 33 commits over the past month, it's clear that the community is actively contributing to this project.
design-image-studio, with a growth score of 43.50 and 72 stars, is another tool gaining attention for its ability to transform vague visual demands into high-quality design images. Its unique approach to image generation has resonated with users, leading to a steady increase in popularity.
ERNIE-Image, an open text-to-image generation model developed by Baidu, boasts a growth score of 39.50 and 291 stars. This tool's state-of-the-art performance among open-weight text-to-image models has made it an attractive option for developers, with 21 commits over the past month demonstrating its active development.
ashim, a Docker container housing 30+ tools for image processing, including resizing, compressing, and removing backgrounds, has seen significant growth with a score of 34.30 and 834 stars. Its appeal lies in its ability to perform complex tasks locally, without relying on cloud services or telemetry, making it an attractive option for users prioritizing data security.
MOSS-VL, the core multimodal model series within the OpenMOSS ecosystem, has achieved a growth score of 19.12 and 225 stars. Its dedication to visual understanding has resonated with researchers and developers, leading to a steady increase in popularity and 37 commits over the past month.
AlayaRenderer, an AI-native renderer for games and virtual worlds, boasts a growth score of 17.36 and 568 stars. This tool's unique approach to rendering has caught the attention of game developers and virtual world creators, who are eager to explore its potential for generating immersive experiences.
forge-film, a multi-model DAG-driven parallel AI film generation engine, has seen growth with a score of 13.83 and 492 stars. Its ability to generate film scenes simultaneously, rather than one by one, has made it an attractive option for filmmakers and content creators looking to streamline their workflow.
Camera-Transformer-1, which understands user intent and generates videos with precise camera control, boasts a growth score of 11.55 and 96 stars. This tool's innovative approach to video generation has piqued the interest of developers and users alike, leading to a steady increase in popularity.
awesome-generative-models, a repository providing weekly real-time updates on diffusion generation models, has seen moderate growth with a score of 8.43 and 100 stars. Its value lies in its curation of cutting-edge research and tools, making it an essential resource for researchers and developers in the field.
Finally, PixelDiT, a pixel diffusion transformer for image generation, boasts a growth score of 3.64 and 126 stars. Although its growth has been slower compared to other repositories on this list, its unique approach to image generation makes it worth keeping an eye on in the future.
JoyAI-Image, a unified multimodal foundation model for image understanding and generation, has seen an impressive growth score of 63.83 and now boasts 1,929 stars. Its popularity can be attributed to its versatility in handling various tasks such as text-to-image generation and instruction-guided image editing. With 33 commits over the past month, it's clear that the community is actively contributing to this project.
design-image-studio, with a growth score of 43.50 and 72 stars, is another tool gaining attention for its ability to transform vague visual demands into high-quality design images. Its unique approach to image generation has resonated with users, leading to a steady increase in popularity.
ERNIE-Image, an open text-to-image generation model developed by Baidu, boasts a growth score of 39.50 and 291 stars. This tool's state-of-the-art performance among open-weight text-to-image models has made it an attractive option for developers, with 21 commits over the past month demonstrating its active development.
ashim, a Docker container housing 30+ tools for image processing, including resizing, compressing, and removing backgrounds, has seen significant growth with a score of 34.30 and 834 stars. Its appeal lies in its ability to perform complex tasks locally, without relying on cloud services or telemetry, making it an attractive option for users prioritizing data security.
MOSS-VL, the core multimodal model series within the OpenMOSS ecosystem, has achieved a growth score of 19.12 and 225 stars. Its dedication to visual understanding has resonated with researchers and developers, leading to a steady increase in popularity and 37 commits over the past month.
AlayaRenderer, an AI-native renderer for games and virtual worlds, boasts a growth score of 17.36 and 568 stars. This tool's unique approach to rendering has caught the attention of game developers and virtual world creators, who are eager to explore its potential for generating immersive experiences.
forge-film, a multi-model DAG-driven parallel AI film generation engine, has seen growth with a score of 13.83 and 492 stars. Its ability to generate film scenes simultaneously, rather than one by one, has made it an attractive option for filmmakers and content creators looking to streamline their workflow.
Camera-Transformer-1, which understands user intent and generates videos with precise camera control, boasts a growth score of 11.55 and 96 stars. This tool's innovative approach to video generation has piqued the interest of developers and users alike, leading to a steady increase in popularity.
awesome-generative-models, a repository providing weekly real-time updates on diffusion generation models, has seen moderate growth with a score of 8.43 and 100 stars. Its value lies in its curation of cutting-edge research and tools, making it an essential resource for researchers and developers in the field.
Finally, PixelDiT, a pixel diffusion transformer for image generation, boasts a growth score of 3.64 and 126 stars. Although its growth has been slower compared to other repositories on this list, its unique approach to image generation makes it worth keeping an eye on in the future.