Today's AI Agent: Fastest-Growing Projects — June 01, 2026
Today's the AI Agent space, there's a notable trend towards more specialized and versatile applications of artificial intelligence, with developers focusing on optimizing performance for smaller models and enhancing user interaction through creative coding techniques. One standout project is Doorman11991/smallcode, an AI coding agent tailored for efficient use with small LLMs, achieving impressive benchmark results with its 4B-active model. With a growth score of 95.25 and over 1,600 stars, it clearly resonates with developers looking to maximize the capabilities of smaller models.
Datawhalechina's Agent-Learning-Hub is another significant project this week, serving as an extensive resource for learning about AI agents through curated routes and materials. Its high growth score of 93.10 alongside over 2,000 stars indicates its popularity among those seeking comprehensive education in the field.
UditAkhourii's ADHD (A Divergent Hierarchical Decision-making) is a unique skill set for coding agents that employs tree-of-thought with pruning techniques to enhance creativity and interdisciplinary problem-solving. With a growth score of 72.00, it stands out due to its innovative approach to AI-driven decision-making processes.
Basketikun's infinite-canvas offers an open-source work platform designed specifically for AI artists and creators, integrating features like AI image generation, video creation, and dialogue assistance. Its integration with OpenAI APIs makes it highly versatile, contributing to a growth score of 64.86 and over 800 stars.
Yapeth's MaaNTE is an automated assistant based on the MAAFramework designed for repetitive tasks in gaming environments. Despite its niche focus, it garners interest from users looking for specialized automation solutions, reflected in its steady increase with a growth score of 58.75 and around 200 stars.
OpenSquilla's opensquilla is an innovative AI agent aiming to achieve higher intelligence density within the same token budget, setting itself apart through efficiency-focused design. With over 2,100 stars and a robust development pace reflected in its growth score of 55.54, it appeals to developers prioritizing performance optimization.
WenyuChiou's awesome-agentic-ai-zh offers a structured learning path for AI agents with multilingual support, making it an invaluable resource for beginners and intermediate learners alike. Its high star count (1,822) and growth score of 54.50 highlight its broad appeal across different language communities.
AmmarkoV's SAM3DBody-cpp is a real-time full-body reconstruction tool that outputs multiperson BVH data using ONNX and ggml frameworks for efficient runtime performance. Its specialized focus on real-time body tracking has earned it a growth score of 53.62 and around 400 stars, indicating interest from developers working with complex motion capture applications.
XingYu-Zhong's DeepSeek-GUI provides an AI agent workspace tailored to DeepSeek models, offering both Code and Claw modes directly within the application interface. Its growing popularity is evident in its growth score of 51.55 and over 700 stars, appealing to developers looking for integrated AI work environments.
Lastly, python-telegramBot's ai-auto-trading showcases a crypto trading bot powered by an LLM agent, facilitating quantitative trading strategies across multiple exchanges like Binance and Gate.io. Its specialized focus on financial applications is attracting attention, as indicated by its growth score of 50.76 and around 200 stars.
These projects highlight the diverse landscape of AI agents, from general learning resources to specialized tools for coding, creative work, and finance, each demonstrating unique strengths that cater to specific developer needs and interests.
Datawhalechina's Agent-Learning-Hub is another significant project this week, serving as an extensive resource for learning about AI agents through curated routes and materials. Its high growth score of 93.10 alongside over 2,000 stars indicates its popularity among those seeking comprehensive education in the field.
UditAkhourii's ADHD (A Divergent Hierarchical Decision-making) is a unique skill set for coding agents that employs tree-of-thought with pruning techniques to enhance creativity and interdisciplinary problem-solving. With a growth score of 72.00, it stands out due to its innovative approach to AI-driven decision-making processes.
Basketikun's infinite-canvas offers an open-source work platform designed specifically for AI artists and creators, integrating features like AI image generation, video creation, and dialogue assistance. Its integration with OpenAI APIs makes it highly versatile, contributing to a growth score of 64.86 and over 800 stars.
Yapeth's MaaNTE is an automated assistant based on the MAAFramework designed for repetitive tasks in gaming environments. Despite its niche focus, it garners interest from users looking for specialized automation solutions, reflected in its steady increase with a growth score of 58.75 and around 200 stars.
OpenSquilla's opensquilla is an innovative AI agent aiming to achieve higher intelligence density within the same token budget, setting itself apart through efficiency-focused design. With over 2,100 stars and a robust development pace reflected in its growth score of 55.54, it appeals to developers prioritizing performance optimization.
WenyuChiou's awesome-agentic-ai-zh offers a structured learning path for AI agents with multilingual support, making it an invaluable resource for beginners and intermediate learners alike. Its high star count (1,822) and growth score of 54.50 highlight its broad appeal across different language communities.
AmmarkoV's SAM3DBody-cpp is a real-time full-body reconstruction tool that outputs multiperson BVH data using ONNX and ggml frameworks for efficient runtime performance. Its specialized focus on real-time body tracking has earned it a growth score of 53.62 and around 400 stars, indicating interest from developers working with complex motion capture applications.
XingYu-Zhong's DeepSeek-GUI provides an AI agent workspace tailored to DeepSeek models, offering both Code and Claw modes directly within the application interface. Its growing popularity is evident in its growth score of 51.55 and over 700 stars, appealing to developers looking for integrated AI work environments.
Lastly, python-telegramBot's ai-auto-trading showcases a crypto trading bot powered by an LLM agent, facilitating quantitative trading strategies across multiple exchanges like Binance and Gate.io. Its specialized focus on financial applications is attracting attention, as indicated by its growth score of 50.76 and around 200 stars.
These projects highlight the diverse landscape of AI agents, from general learning resources to specialized tools for coding, creative work, and finance, each demonstrating unique strengths that cater to specific developer needs and interests.