Skip to content

HippoRAG

A RAG framework inspired by human long-term memory

Visit official site

What HippoRAG is

The framework is described as a retrieval-augmented generation approach inspired by human long-term memory, letting large language models integrate knowledge across external documents.

What you can do with it

  • Keep integrating external document knowledge
  • Use predefined queries and API-based retrieval
  • Combine with IRCoT for complementary performance

Getting started

  • Create a conda environment and install dependencies
  • Prepare the retrieval corpus and query files in the required format
  • Connect online or offline LLMs through LangChain

Pros & cons

✓ What we like

  • Setup and usage guides are provided
  • Several retrieval strategies are offered

! What to watch out for

  • A conda environment and dependencies are required
  • The corpus and query files must follow a set format

FAQ

What is the design inspired by?

Human long-term memory, following neurobiological principles.

How are LLMs connected?

Through LangChain, using online LLM APIs or offline LLM deployments.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review