Today's Image & Video Generation: Fastest-Growing Projects — April 15, 2026
Today's Image & Video Generation, we're seeing a surge in open-source tools that leverage AI to generate high-quality visuals. The trend is shifting towards more specialized models that can perform specific tasks, such as text-to-image generation and video production. With the increasing popularity of these tools, developers are flocking to GitHub to explore and contribute to these projects.
ERNIE-Image, developed by Baidu, has seen significant growth with a score of 95.00 and 70 stars. This open-source text-to-image generation model uses a single-stream Diffusion Transformer (DiT) to achieve state-of-the-art performance among open-weight models. Its impressive capabilities have drawn attention from developers looking for efficient image generation solutions.
OpenMontage, on the other hand, has garnered an impressive 1,680 stars and a growth score of 91.94. This world's first open-source agentic video production system allows users to turn their AI coding assistant into a full video production studio with its 11 pipelines, 49 tools, and 400+ agent skills. Its massive community engagement is likely due to its comprehensive features and potential for creative applications.
JoyAI-Image from jd-opensource boasts a growth score of 71.73 and 1,681 stars. As the unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing, JoyAI-Image offers a versatile toolset for developers working with visual data. Its broad range of capabilities has attracted significant attention from the developer community.
Ashim-hq's ashim project has seen notable growth with a score of 39.91 and 706 stars. This Docker container-based tool provides over 30 image processing tools, including resizing, compression, background removal, and more, all without relying on cloud services or telemetry. Its focus on local AI and user data security is likely driving its popularity.
MOSS-VL from OpenMOSS has achieved a growth score of 31.79 and 211 stars. As the core multimodal model series for visual understanding within the OpenMOSS ecosystem, MOSS-VL offers a robust foundation for developers working with visual data. Its specialized focus on visual understanding is likely attracting attention from researchers and developers in related fields.
Modly by lightningpixel has gained traction with a growth score of 29.64 and 852 stars. This desktop app uses local AI to generate 3D models from images, running entirely on the user's GPU. Its ability to leverage local computing resources is likely driving its popularity among developers looking for efficient 3D modeling solutions.
Other notable projects in this space include AlayaRenderer by ShandaAI, forge-film by F-R-L, Pilipili-AutoVideo by OpenDemon, and awesome-generative-models by jiangchaokang. While these tools have seen significant growth, their descriptions are less detailed or lack clear explanations of their functionality, making it harder to understand the reasons behind their popularity.
ERNIE-Image, developed by Baidu, has seen significant growth with a score of 95.00 and 70 stars. This open-source text-to-image generation model uses a single-stream Diffusion Transformer (DiT) to achieve state-of-the-art performance among open-weight models. Its impressive capabilities have drawn attention from developers looking for efficient image generation solutions.
OpenMontage, on the other hand, has garnered an impressive 1,680 stars and a growth score of 91.94. This world's first open-source agentic video production system allows users to turn their AI coding assistant into a full video production studio with its 11 pipelines, 49 tools, and 400+ agent skills. Its massive community engagement is likely due to its comprehensive features and potential for creative applications.
JoyAI-Image from jd-opensource boasts a growth score of 71.73 and 1,681 stars. As the unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing, JoyAI-Image offers a versatile toolset for developers working with visual data. Its broad range of capabilities has attracted significant attention from the developer community.
Ashim-hq's ashim project has seen notable growth with a score of 39.91 and 706 stars. This Docker container-based tool provides over 30 image processing tools, including resizing, compression, background removal, and more, all without relying on cloud services or telemetry. Its focus on local AI and user data security is likely driving its popularity.
MOSS-VL from OpenMOSS has achieved a growth score of 31.79 and 211 stars. As the core multimodal model series for visual understanding within the OpenMOSS ecosystem, MOSS-VL offers a robust foundation for developers working with visual data. Its specialized focus on visual understanding is likely attracting attention from researchers and developers in related fields.
Modly by lightningpixel has gained traction with a growth score of 29.64 and 852 stars. This desktop app uses local AI to generate 3D models from images, running entirely on the user's GPU. Its ability to leverage local computing resources is likely driving its popularity among developers looking for efficient 3D modeling solutions.
Other notable projects in this space include AlayaRenderer by ShandaAI, forge-film by F-R-L, Pilipili-AutoVideo by OpenDemon, and awesome-generative-models by jiangchaokang. While these tools have seen significant growth, their descriptions are less detailed or lack clear explanations of their functionality, making it harder to understand the reasons behind their popularity.