unstract
LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
The Unstract MCP server enables AI agents to extract structured JSON data from unstructured documents using large language models. Built by Zipstack, this server integrates document processing capabilities into agent workflows, allowing assistants like Claude to transform PDFs, images, and scans into structured data through natural language prompts. It bridges Unstract's extraction platform with the Model Context Protocol ecosystem, making document intelligence accessible to conversational AI systems. The server works alongside Unstract's broader platform features including API deployment and ETL pipelines.
Key Features
Use Cases
- 01Extract structured data from financial statements and invoices for automated bookkeeping
- 02Process insurance claims documents and policy forms for compliance workflows
- 03Convert medical records and healthcare documents into database-ready JSON
- 04Automate KYC and compliance document verification with structured field extraction
- 05Transform income tax forms and supporting documents into queryable data
- 06Build document processing agents that handle vendor-specific formats without custom templates
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
unstract — FAQ
What is the Unstract MCP server?+
The Unstract MCP server is a Model Context Protocol integration that allows AI agents like Claude to extract structured JSON from unstructured documents (PDFs, images, scans) using LLMs. It exposes Unstract's document processing capabilities through the MCP standard.
How do I install and run Unstract locally?+
Clone the repository, ensure you have Docker and Docker Compose installed with at least 8GB RAM, then run ./run-platform.sh. Access the platform at http://frontend.unstract.localhost with default credentials (username: unstract, password: unstract).
Which AI clients support the Unstract MCP server?+
The Unstract MCP server works with Claude and other Model Context Protocol-compatible clients. The documentation specifically mentions integration with Claude agents.
Do I need API keys to use Unstract?+
You'll need API keys for whichever LLM provider you choose (OpenAI, Anthropic, AWS Bedrock, etc.). Unstract also supports local models through Ollama that don't require external API keys. The platform itself is open source and runs locally.
Is Unstract free to use?+
Yes, Unstract is released under the AGPL-3.0 open source license and can be run locally for free. Enterprise features like SSO, compliance certifications, and managed cloud hosting are available through paid plans.
What should I do about the encryption key?+
After installation, immediately copy the ENCRYPTION_KEY value from backend/.env or platform-service/.env to a secure location. This key encrypts adapter credentials, and losing it makes existing adapters permanently inaccessible.
How do I install unstract?+
Open the source repository on GitHub and follow its README. unstract is a mcp server — MCP Agents Market links you directly to the official repo.
Is unstract free?+
unstract is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.