sifter-ai/sifter

Overview

Sifter is an open-source, developer-first document extraction engine. It uses TypeScript to convert unstructured documents such as invoices, contracts, receipts, and reports into a structured, queryable database. The extraction process is designed to be self-hosted and integrated into developer workflows.

Best for

Best for
Developers needing an open-source document extraction engine to build structured databases from unstructured files

Use cases

• Extracting structured data from invoices and receipts
• Parsing contracts into queryable records for analysis
• Building a searchable database from reports and documents

How to use

Install

pip install sifter-ai

Tools exposed

• uvx

Tested with

Claude Desktop, Cursor

Example client config

{\n  "mcpServers": {\n    "sifter": {\n      "command": "uvx",\n      "args": ["sifter-mcp", "--base-url", "http://localhost:8000"],\n      "env": { "SIFTER_API_KEY": "sk-dev" }\n    }\n  }\n}

Notes

Sifter is an open-source, developer-first document extraction engine. It uses TypeScript to convert unstructured documents such as invoices, contracts, receipts, and reports into a structured, queryable database. The extraction process is designed to be self-hosted and integrated into developer workflows.

44 stars on GitHub. Last updated 2026-05-29. Licensed MIT.

Use cases

• Extracting structured data from invoices and receipts
• Parsing contracts into queryable records for analysis
• Building a searchable database from reports and documents

Pros

• Open-source and self-hostable, giving full control over data
• Developer-first design with a TypeScript API for easy integration
• Turns unstructured documents into structured, queryable data

Cons

• Small community (44 stars) may limit support and contributions
• May require significant setup for complex or varied document types
• Not optimized for real-time extraction at scale without additional tuning

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.