yifanfeng97/Hyper-Extract

Hypergraph is more powerful. Transform unstructured text into structured knowledge with LLMs. Graphs, hypergraphs, and spatio-temporal extractions — with one command.

What it solves

Hyper-Extract is a CLI tool and framework designed to transform unstructured text documents into structured, strongly-typed "Knowledge Abstracts." It eliminates the manual effort of reading and organizing information by automatically extracting data into various formats, from simple lists to complex spatio-temporal graphs, allowing users to query and visualize the resulting knowledge base.

How it works

The system uses a three-layer architecture consisting of Auto-Types (8 strongly-typed data structures), Methods (extraction algorithms like GraphRAG and LightRAG), and Templates (80+ domain-specific presets). It leverages LLMs with function-calling capabilities (such as GPT-4o, Claude, and DeepSeek) and OpenAI-compatible embedding models to parse documents based on a selected template. The extracted data is stored as a persistent knowledge base that can be incrementally evolved as new documents are added.

Who it’s for

  • Researchers who need to turn academic papers into interactive concept graphs.
  • Financial Analysts who want to extract entities and metrics from earnings reports.
  • Developers looking for a local, private deployment using vLLM to keep data on-premise.
  • Knowledge Managers who use Obsidian to organize their notes via wikilinks.

Highlights

  • Diverse Knowledge Structures: Supports 8 types, including Hypergraphs and Spatio-Temporal Graphs.
  • Extensive Template Library: Over 80 YAML templates for domains like Finance, Legal, and Medical.
  • Multi-Model Support: Compatible with OpenAI, Anthropic, DeepSeek, Alibaba Cloud, and local vLLM deployments.
  • Obsidian Integration: Ability to export extracted graphs directly into an Obsidian vault.
  • MCP Server: Exposes knowledge abstracts to MCP-capable assistants like Claude Desktop and IDE agents.

Related

  • Project
    shcherbak-ai/contextgem

    An open-source LLM framework for extracting structured data and insights from documents with precise source references and automatic justifications.

  • Project
    docling-project/docling-graph

    A toolkit that transforms unstructured documents into validated Pydantic objects and directed knowledge graphs for high-precision entity relationship extraction.

  • Project
    google/langextract

    A Python library that uses LLMs to extract structured, grounded information from unstructured text, featuring optimized processing for long documents and interactive visualization.

  • Project
    Unstructured-IO/unstructured

    An open-source library for ingesting and pre-processing unstructured data from over 60 file types into structured formats optimized for LLMs and vector databases.

  • Project
    emcf/thepipe

    A multimodal data extraction package that scrapes clean markdown and structured data from complex documents using vision-language models for RAG and LLM applications.