pdf-reader-mcp
Give your AI agent eyes for PDFs — structured text, tables, OCR, visual evidence, and page-level citations via MCP. Native Rust, local-first.
Citra is a local-first MCP server that gives AI agents the ability to read, extract, and cite content from PDF documents with structured evidence. Built with a native Rust engine, it provides text extraction, table parsing, OCR for scanned documents, and page-level citations with bounding-box coordinates. The server runs entirely on your machine without requiring API keys or cloud services, and can be launched instantly using npx. It returns structured evidence that agents can reference with specific page numbers, table cells, and geometric coordinates instead of hallucinating or guessing PDF content.
Key Features
Use Cases
- 01Extracting financial data from quarterly reports with table cells cited by page and row/column coordinates
- 02Analyzing research papers with preserved heading structure, reading order, and precise page-level quotations
- 03Processing scanned documents and invoices using OCR with evidence trails linking extracted text to source pages
- 04Building citation-capable document assistants that can defend their answers with specific page numbers and geometric locations
- 05Automating contract review with table structure preservation and verifiable text extraction
- 06Creating research tools that extract structured bibliographic data and quotes from academic PDFs
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.
pdf-reader-mcp — FAQ
What is Citra?+
Citra (also known as pdf-reader-mcp) is an MCP server that enables AI agents to read PDF files with structured evidence including text, tables, OCR, and page-level citations. It runs locally using a native Rust engine and provides agents with geometric coordinates and provenance data they can cite instead of hallucinating content.
How do I install the Citra MCP server?+
You can run Citra instantly with npx -y @sylphx/citra for any MCP client, or add it to Claude Desktop/Cursor/VS Code by putting the npx command and args in your MCP configuration JSON. For global installation, use npm i -g @sylphx/citra to get the citra command.
Which AI clients work with this MCP server?+
Citra works with Claude Desktop, Claude Code, Cursor, VS Code, Codex, and any other client that supports the MCP stdio protocol. It also supports dual-era hosts that send server/discover before initialize, such as the Gemini Antigravity CLI.
Does Citra require API keys or internet access?+
No, Citra is completely local-first and requires no API keys or cloud services. PDFs are processed entirely on your machine using the native Rust engine, and no document data is sent to external servers.
Is Citra free to use?+
Yes, Citra is open source and released under the MIT License, making it free for both personal and commercial use.
What platforms does the native Rust engine support?+
The native engine supports macOS (arm64 and x64), Linux (x64 and arm64 GNU), and Windows (x64 MSVC). The appropriate platform-specific package is automatically selected during installation for your operating system.
How do I install pdf-reader-mcp?+
Open the source repository on GitHub and follow its README. pdf-reader-mcp is a mcp server — MCP Agents Market links you directly to the official repo.
Is pdf-reader-mcp free?+
pdf-reader-mcp is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.