RAG and document tools for agent memory
Retrieval tools that let agents search documents, PDFs and knowledge bases and answer from them (RAG). Of the 105 tools on our memory for agents page, 25 fit this narrower job. By pricing: 20 open source, 1 source-available, 2 free or freemium, 1 paid, usage-based or enterprise, 1 pricing on the maker's site. The most-starred open repos here are MarkItDown (190,824 stars), RAGFlow (92,335 stars), Headroom (75,629 stars). Most come from the memory & RAG (22), frameworks (2), productivity (1) categories.
Updated
MarkItDown
Microsoft's Python tool and MCP server that converts PDFs, Office documents, images, audio and web pages to Markdown for LLM and agent pipelines.
Added
RAGFlow
Open-source RAG engine with deep document parsing and agent features for building grounded question-answering over your files.
Added
Headroom
Library, proxy and MCP server that compresses tool outputs, logs, files and RAG chunks locally before they reach the LLM, cutting tokens for coding agents.
Added
Mem0
Mem0 is an open-source memory layer for AI agents that extracts facts from conversations in a single add-only pass, links entities across memories and retrieves them in later sessions, with an Apache-2.0 SDK and a managed platform.
Added
PageIndex
Document index for reasoning-based, vectorless RAG that lets agents navigate long documents like a table of contents.
Added
Chroma
Open-source search and vector database widely used for agent memory and RAG.
Added
llmware
Framework for building private, local RAG and agent applications with small specialized models on AI PCs, edge and self-hosted setups.
Added
txtai
All-in-one Python framework for semantic search, LLM orchestration and agents built on an embeddings database of vector, graph and relational indexes.
Added
MemOS
Memory operating system for LLMs and agents with persistent memory and retrieval.
Added
LanceDB
Open-source multimodal vector database for search, RAG and agent memory.
Added
xerj
A local-first search engine and MCP server for agents: one Rust binary indexes code, docs, logs, CSV/JSONL and PDFs for full-text, vector and hybrid search plus agent memory, behind an Elasticsearch-compatible API.
Added
Activeloop
Deep Lake, a multimodal database for AI that stores vectors, images, text and video so agents and RAG apps can search and retrieve them.
Added
Zep
Context layer that unifies business data, documents and conversations into memory so agents complete tasks correctly.
Added
OpenKnowledge
OpenKnowledge is a free, local-first markdown editor and knowledge base built for working alongside AI agents. It looks and feels like a Notion-style WYSIWYG editor, with callouts, tabs, Mermaid diagrams and media blocks, but every page stays a plain markdown file on your disk. Coding agents such as Claude Code, Codex, Cursor, Gemini CLI and dozens of others can read and write the same files through its integrations, so specs, team wikis and LLM wikis stay in one place that both people and agents keep up to date. It ships as a desktop app plus a web UI and CLI.
Added
Memgraph
In-memory graph database used as the graph layer for GraphRAG pipelines and agent memory, with Cypher queries plus vector and text search.
Added
SimpleMem
Research-backed lifelong memory library for LLM agents that stores, compresses and retrieves text and multimodal memories, usable over MCP.
Added
Knowhere
Document parsing and retrieval system that turns messy files into structured, navigable memory for AI agents and RAG.
Added
seekdb
AI-native search database for agent storage combining vector, text and structured data.
Added
Neuron AI
PHP framework for creating and orchestrating AI agents with LLMs, tools, vector stores and memory.
Added
Agentset
Open-source RAG platform for building agents and apps that answer accurately from your documents.
Added
Pixeltable
Multimodal data layer that combines storage, transforms, indexing and retrieval in one Python library for building agents over images, video and documents.
Added
Qdrant MCP Server
Official Qdrant MCP server that lets agents store and retrieve memories and documents in the Qdrant vector database.
Added
BaseAI
Web AI framework from Langbase for building AI agents with memory (RAG) and deploying them serverless.
Added
Extend MCP
Extend's MCP server for turning documents into structured data: parse, extract, classify, split and fill PDF forms.
Added
LlamaParse
LlamaIndex's document processing platform that parses PDFs, scans and Office files into markdown or JSON and extracts, classifies and indexes them for agents.
Added
GitHub repos (23)
| Repo | GitHub stars | Licence | Last push |
|---|---|---|---|
| MarkItDown ↗ | 190,824 | MIT | 2026-10-20 |
| RAGFlow ↗ | 92,335 | Apache License 2.0 | 2026-10-21 |
| Headroom ↗ | 75,629 | Apache License 2.0 | 2026-10-21 |
| Mem0 ↗ | 67,147 | Apache License 2.0 | 2026-10-21 |
| PageIndex ↗ | 39,186 | MIT | 2026-10-21 |
| Chroma ↗ | 29,512 | Apache License 2.0 | 2026-10-17 |
| llmware ↗ | 14,814 | Apache License 2.0 | 2026-10-01 |
| txtai ↗ | 13,024 | Apache License 2.0 | 2026-10-20 |
| MemOS ↗ | 11,892 | Apache License 2.0 | 2026-09-22 |
| LanceDB ↗ | 11,684 | Apache License 2.0 | 2026-10-21 |
| xerj ↗ | 11,507 | — | 2026-10-21 |
| Activeloop ↗ | 9,252 | Apache License 2.0 | 2026-05-21 |
| Zep ↗ | 4,966 | — | 2026-10-20 |
| OpenKnowledge ↗ | 4,646 | — | — |
| Memgraph ↗ | 4,622 | — | 2026-10-21 |
Top 15 of 23 by stars. Stars and licence come from each repo, checked when the listing was last verified.
FAQ
What are the best AI tools for memory for agents focused on RAG and documents?
We list 25. The first three by GitHub stars and featured status are MarkItDown: Microsoft's Python tool and MCP server that converts PDFs, Office documents, images, audio and web pages to Markdown for LLM and agent pipelines.; RAGFlow: Open-source RAG engine with deep document parsing and agent features for building grounded question-answering over your files.; Headroom: Library, proxy and MCP server that compresses tool outputs, logs, files and RAG chunks locally before they reach the LLM, cutting tokens for coding agents.
Are there free or open-source options?
Yes: 20 open source, 1 source-available, 2 free or freemium, 1 paid, usage-based or enterprise, 1 pricing on the maker's site. Pricing is as each maker states it.
More from memory for agents
Tools are grouped by what each says it does. Missing one? Submit it free.