HippoRAG

HippoRAG is a memory framework for large language models that recognizes and uses connections in new knowledge.

HippoRAG is a memory framework designed for large language models that mimics human long-term memory. It allows models to continuously integrate knowledge across external documents by recognizing and utilizing connections in new data. The system improves associativity and sense-making in advanced retrieval-augmented generation systems while maintaining cost and latency efficiency. It supports local deployment via vLLM, OpenAI, Azure OpenAI, and multiple vector store backends like Qdrant, ChromaDB, and Milvus. Users can interact with the system through a Python API or an MCP server for agents and clients like Claude Code, Cursor, and VS Code.

HippoRAG at a glance

Best for: Software engineer, Researcher
Runs on: Python, API, CLI, self-hosted, open source
Pricing: Free
Released: May 2024
Category: AI research tool
Website: github.com/OSU-NLP-Group/HippoRAG

Key features

Integrates knowledge across external documents
Supports multi-hop retrieval and sense-making
Provides a Model Context Protocol server
Connects to Qdrant, ChromaDB, and Milvus
Deploys locally with vLLM support
Compatible with OpenAI and Azure OpenAI

Use it for

Retrieve relevant passages for a query with scores
Answer questions grounded in retrieved passages
Add documents and manage a persistent index
Run continual learning and reproducibility experiments

Alternatives to HippoRAG

LightRAG — It uses a dual-layer architecture to reduce LLM calls and retrieval latency.
FlashRAG — It provides pre-processed benchmark datasets and pre-implemented algorithms.
OpenVLHarness — Lifts six multimodal model backbones by 9.7 to 12.6 points across 22 benchmarks.
UltraRAG — Build complex RAG workflows using modular MCP components and YAML configuration.
WeKnora — Combines RAG, multi-step agents, and automatic Wiki generation in one platform.
Hemory — Connects real-world conversations to your AI agents via MCP.
Morphik — Provides native Model Context Protocol support and granular multi-user access.
GraphRAG — Enhances LLM reasoning over private data using knowledge graph structures.
HyperResearch — Generates verified reports with citations checked against full source texts and a persistent vault.
LEANN — Efficiently index millions of documents locally with 97% storage savings.
My RAG — Compare Python and Java implementations sharing identical data.
Ncm — Generates text significantly faster than diffusion baselines while maintaining high quality.

All HippoRAG alternatives