Today's RAG & Vector Databases: Fastest-Growing Projects — May 25, 2026
Today's the RAG & Vector Databases space, there's a noticeable trend towards enterprise-focused solutions that integrate large language models (LLMs) and retrieval-augmented generation techniques into knowledge management systems. Additionally, there is growing interest in tools designed to assist developers with learning and implementing RAG applications. Leading Today's list is nduckmink/arkon, which has seen significant growth and engagement.
Arkon, developed by nduckmink, serves as an enterprise AI knowledge hub that allows teams to self-host a knowledge base for managing retrieval-augmented generation contexts and access policies. The tool supports the Model Context Protocol (MCP) for connecting with LLMs like Claude, enabling automated and secure integration of organizational knowledge. With its growth score of 40.70 and over 859 stars on GitHub, Arkon's popularity is likely due to its comprehensive approach to managing enterprise AI contexts.
ZJunCher/xiaoyan-ai-dev-assistant is another notable project this week. It combines RAG with multi-hop reasoning to offer a development assistant that supports team knowledge sharing and is also suitable for newcomers learning how to develop RAG applications. With 27.62 growth score and 106 stars, xiaoyan-ai-dev-assistant appears to be growing due to its dual purpose of aiding both experienced developers and beginners in the realm of RAG.
GasolSun36/PyRAG is a lesser but still noteworthy project that focuses on executable multi-hop reasoning for retrieval-augmented generation. The tool aims to provide cheap, efficient retrieval methods for complex reasoning tasks. Despite having a lower growth score of 2.38 and fewer stars (24), PyRAG maintains steady development with 5 commits in the last month, indicating continued interest from its core user base who are interested in advanced RAG techniques.
These projects highlight the diverse applications of RAG technologies across both enterprise knowledge management and developer education, reflecting a growing ecosystem around these tools.
Arkon, developed by nduckmink, serves as an enterprise AI knowledge hub that allows teams to self-host a knowledge base for managing retrieval-augmented generation contexts and access policies. The tool supports the Model Context Protocol (MCP) for connecting with LLMs like Claude, enabling automated and secure integration of organizational knowledge. With its growth score of 40.70 and over 859 stars on GitHub, Arkon's popularity is likely due to its comprehensive approach to managing enterprise AI contexts.
ZJunCher/xiaoyan-ai-dev-assistant is another notable project this week. It combines RAG with multi-hop reasoning to offer a development assistant that supports team knowledge sharing and is also suitable for newcomers learning how to develop RAG applications. With 27.62 growth score and 106 stars, xiaoyan-ai-dev-assistant appears to be growing due to its dual purpose of aiding both experienced developers and beginners in the realm of RAG.
GasolSun36/PyRAG is a lesser but still noteworthy project that focuses on executable multi-hop reasoning for retrieval-augmented generation. The tool aims to provide cheap, efficient retrieval methods for complex reasoning tasks. Despite having a lower growth score of 2.38 and fewer stars (24), PyRAG maintains steady development with 5 commits in the last month, indicating continued interest from its core user base who are interested in advanced RAG techniques.
These projects highlight the diverse applications of RAG technologies across both enterprise knowledge management and developer education, reflecting a growing ecosystem around these tools.