Today's RAG & Vector Databases: Fastest-Growing Projects — May 02, 2026
Today's the RAG & Vector Databases space, we're seeing a surge in innovative tools that leverage knowledge graphs and hybrid search capabilities to compound information over time. Local-first RAG knowledge base compilers are gaining traction, while agentic memory solutions for CTI are also making waves.
Swarmvault, with a growth score of 19.21 and 306 stars, is an example of the former trend. This tool allows users to turn raw research into a persistent markdown wiki, knowledge graph, and hybrid search that compounds over time, leveraging MCP servers for Claude Code, Codex, OpenCode, and OpenClaw. Its rapid growth can be attributed to its unique ability to organize and interlink knowledge bases in a local-first approach.
Zettelforge, boasting a growth score of 12.52 and 33 stars, takes a different approach with its agentic memory solution for CTI in Python. By integrating STIX knowledge graphs and threat-actor alias resolution, Zettelforge provides offline-first RAG capabilities and an MCP server for Claude Code and LangChain agents. Its popularity stems from its ability to enhance security analysts' threat intelligence analysis capabilities.
On the other hand, LLM Wiki is a cross-platform desktop application that turns documents into an organized knowledge base automatically, with a growth score of 7.09 and an impressive 5,421 stars. Instead of traditional RAG, LLM incrementally builds and maintains a persistent wiki from sources, making it a favorite among users seeking to streamline their knowledge management processes.
CTI-RAG, despite having zero commits in the past month, still holds a respectable growth score of 5.93 and 201 stars. This Retrieval-Augmented Generation framework integrates knowledge graph and causal reasoning capabilities for CTI analysis, providing security analysts with an intelligent threat intelligence tool. Its stagnant commit activity notwithstanding, its unique feature set continues to attract interest.
Lastly, Zero-to-AI-Fullstack, a learning project by a Java backend engineer, rounds out our list with a growth score of 4.31 and 150 stars. By documenting their journey in learning AI full-stack development using Python, FastAPI, RAG, pgvector, and Next.js, the author is inadvertently creating a valuable resource for others seeking to follow in their footsteps. While not directly related to RAG & Vector Databases, its inclusion showcases the growing interest in these technologies.
Overall, Today's RAG & Vector Databases landscape is marked by innovative solutions that improve knowledge management and threat intelligence analysis capabilities. As users increasingly seek out tools that can efficiently organize and compound information over time, we expect these trends to continue shaping the space in the weeks to come.
Swarmvault, with a growth score of 19.21 and 306 stars, is an example of the former trend. This tool allows users to turn raw research into a persistent markdown wiki, knowledge graph, and hybrid search that compounds over time, leveraging MCP servers for Claude Code, Codex, OpenCode, and OpenClaw. Its rapid growth can be attributed to its unique ability to organize and interlink knowledge bases in a local-first approach.
Zettelforge, boasting a growth score of 12.52 and 33 stars, takes a different approach with its agentic memory solution for CTI in Python. By integrating STIX knowledge graphs and threat-actor alias resolution, Zettelforge provides offline-first RAG capabilities and an MCP server for Claude Code and LangChain agents. Its popularity stems from its ability to enhance security analysts' threat intelligence analysis capabilities.
On the other hand, LLM Wiki is a cross-platform desktop application that turns documents into an organized knowledge base automatically, with a growth score of 7.09 and an impressive 5,421 stars. Instead of traditional RAG, LLM incrementally builds and maintains a persistent wiki from sources, making it a favorite among users seeking to streamline their knowledge management processes.
CTI-RAG, despite having zero commits in the past month, still holds a respectable growth score of 5.93 and 201 stars. This Retrieval-Augmented Generation framework integrates knowledge graph and causal reasoning capabilities for CTI analysis, providing security analysts with an intelligent threat intelligence tool. Its stagnant commit activity notwithstanding, its unique feature set continues to attract interest.
Lastly, Zero-to-AI-Fullstack, a learning project by a Java backend engineer, rounds out our list with a growth score of 4.31 and 150 stars. By documenting their journey in learning AI full-stack development using Python, FastAPI, RAG, pgvector, and Next.js, the author is inadvertently creating a valuable resource for others seeking to follow in their footsteps. While not directly related to RAG & Vector Databases, its inclusion showcases the growing interest in these technologies.
Overall, Today's RAG & Vector Databases landscape is marked by innovative solutions that improve knowledge management and threat intelligence analysis capabilities. As users increasingly seek out tools that can efficiently organize and compound information over time, we expect these trends to continue shaping the space in the weeks to come.