Today's RAG & Vector Databases: Fastest-Growing Projects — April 17, 2026
Today's trends in the RAG & Vector Databases space highlight the increasing focus on retrieval-augmented generation (RAG) tools and their applications in various domains. We see a surge in open-source projects leveraging RAG to improve document search, threat intelligence analysis, and knowledge base management. The growth scores of these repositories reflect the community's interest in exploring the potential of RAG.
joungminsung/OpenDocuments stands out with an impressive growth score of 16.64 and 66 stars, showcasing its popularity among developers. This open-source RAG tool enables AI-powered document search by connecting various platforms like GitHub, Notion, and Google Drive, allowing users to ask questions and receive cited answers. Its self-hosted capabilities using Ollama/OpenAI/Claude have likely contributed to its rapid growth.
yanhua1010/zero-to-ai-fullstack has gained significant traction with a growth score of 14.00 and 147 stars, as it provides a comprehensive learning resource for Java backend engineers transitioning into AI full-stack development. The project's focus on Python, FastAPI, RAG, pgvector, and Next.js has resonated with the community, making it an attractive repository for developers seeking to upskill.
vixhal-baraiya/pageindex-rag, with a growth score of 5.45 and 84 stars, offers a vectorless approach to retrieval-augmented generation (RAG). Its unique methodology has garnered attention from researchers and developers looking for alternative solutions in the RAG space.
Ais1on/CTI-RAG boasts a dedicated following with 46 stars, despite a lower growth score of 4.83. This Retrieval-Augmented Generation framework is specifically designed for Cyber Threat Intelligence (CTI), integrating knowledge graph and causal reasoning capabilities to provide security analysts with advanced threat intelligence analysis tools.
nashsu/llm_wiki has become a sensation in the RAG community, amassing an impressive 1,525 stars and a growth score of 4.61. This cross-platform desktop application transforms documents into organized, interlinked knowledge bases using incremental learning methods, offering a more efficient alternative to traditional RAG approaches.
Vbj1808/Dokis has attracted attention with its lightweight RAG provenance middleware, which verifies claims in LLM responses without requiring an additional LLM call. With a growth score of 2.38 and 34 stars, this tool is slowly gaining recognition for its potential to enhance the reliability of RAG systems.
Lastly, McKern3l/RAGdrag offers a security testing toolkit specifically designed for RAG pipelines, featuring 27 techniques across six kill chain phases mapped to MITRE ATLAS. Despite a lower growth score of 1.83 and 25 stars, this project's focus on pipeline security has resonated with developers seeking to address potential vulnerabilities in their RAG systems.
These repositories demonstrate the ongoing innovation in the RAG & Vector Databases space, as developers continue to explore new applications and improvements for retrieval-augmented generation tools.
joungminsung/OpenDocuments stands out with an impressive growth score of 16.64 and 66 stars, showcasing its popularity among developers. This open-source RAG tool enables AI-powered document search by connecting various platforms like GitHub, Notion, and Google Drive, allowing users to ask questions and receive cited answers. Its self-hosted capabilities using Ollama/OpenAI/Claude have likely contributed to its rapid growth.
yanhua1010/zero-to-ai-fullstack has gained significant traction with a growth score of 14.00 and 147 stars, as it provides a comprehensive learning resource for Java backend engineers transitioning into AI full-stack development. The project's focus on Python, FastAPI, RAG, pgvector, and Next.js has resonated with the community, making it an attractive repository for developers seeking to upskill.
vixhal-baraiya/pageindex-rag, with a growth score of 5.45 and 84 stars, offers a vectorless approach to retrieval-augmented generation (RAG). Its unique methodology has garnered attention from researchers and developers looking for alternative solutions in the RAG space.
Ais1on/CTI-RAG boasts a dedicated following with 46 stars, despite a lower growth score of 4.83. This Retrieval-Augmented Generation framework is specifically designed for Cyber Threat Intelligence (CTI), integrating knowledge graph and causal reasoning capabilities to provide security analysts with advanced threat intelligence analysis tools.
nashsu/llm_wiki has become a sensation in the RAG community, amassing an impressive 1,525 stars and a growth score of 4.61. This cross-platform desktop application transforms documents into organized, interlinked knowledge bases using incremental learning methods, offering a more efficient alternative to traditional RAG approaches.
Vbj1808/Dokis has attracted attention with its lightweight RAG provenance middleware, which verifies claims in LLM responses without requiring an additional LLM call. With a growth score of 2.38 and 34 stars, this tool is slowly gaining recognition for its potential to enhance the reliability of RAG systems.
Lastly, McKern3l/RAGdrag offers a security testing toolkit specifically designed for RAG pipelines, featuring 27 techniques across six kill chain phases mapped to MITRE ATLAS. Despite a lower growth score of 1.83 and 25 stars, this project's focus on pipeline security has resonated with developers seeking to address potential vulnerabilities in their RAG systems.
These repositories demonstrate the ongoing innovation in the RAG & Vector Databases space, as developers continue to explore new applications and improvements for retrieval-augmented generation tools.