• Picks
  • Discover
  • Alternatives
  • About

© 2026 Willet

  • Newsletter
  • About
  • FAQ
  • Terms
  • Privacy & Cookies
Join the Discord (opens in a new tab)
Zipstack avatar

Unstract

Self-hosted platform for extracting structured JSON from documents with LLMs

  • Agent frameworks and tools
  • Data and spreadsheets
  • Self-hosted alternatives
View on GitHub
Website

What it is

Unstract pulls structured JSON out of PDFs, scans, images, office files and spreadsheets by having you describe the fields you want in natural-language prompts. Schemas are built in a web-based Prompt Studio, then deployed as a REST API endpoint or as an ETL pipeline that reads from folders and loads results into a data warehouse. Data teams in finance, insurance, healthcare and compliance run it on their own servers with Docker, plugging in providers such as OpenAI, Anthropic, Bedrock or Ollama, and can also reach it through an MCP server or an n8n node.

Images

Facts

Stars
7,296 GitHub stars.
7,296

Forks
720
Commits
1,772
Contributors
30
Language
Python
License
AGPL-3.0
Last commit
Oct 20, 2026
On GitHub since
February 2024
First release
February 2024
Platforms
web, macOS, Linux, self-hosted
Self-hostable
Yes
Alternative to
Amazon Textract, Google Document AI
Tags
  • mcp-server
  • data-pipeline
  • ocr
  • structured-output
  • document-parsing
  • prompt-engineering

Last synced from GitHub on Oct 20, 2026

The 10 best overlooked open-source projects, every Monday.

Share and more

Comments

No comments yet. Start the conversation.

Similar projects

  • Google avatar

    LangExtract

    Python library for pulling structured, source-linked data out of text with LLMs

    39k