A document processing service using the Docling-MCP library and MCP (Message Control Protocol) for tool integration.
Note
This is an unstable draft implementation which will quickly evolve.
Docling MCP is a service that provides tools for document conversion, processing and generation. It uses the Docling library to convert PDF documents into structured formats and provides a caching mechanism to improve performance. The service exposes functionality through a set of tools that can be called by client applications.
- conversion tools:
- PDF document conversion to structured JSON format (DoclingDocument)
- generation tools:
- Document generation in DoclingDocument, which can be exported to multiple formats
- Local document caching for improved performance
- Support for local files and URLs as document sources
- Memory management for handling large documents
- Logging system for debugging and monitoring
- Milvus upload and retrieval
Install dependencies
uv syncInstall the docling_mcp package
uv pip install -e .After installing the dependencies (uv sync), you can expose the tools of Docling by running
uv run docling-mcp-serverOne of the easiest ways to experiment with the tools provided by Docling-MCP is to leverage
To enable Claude for Desktop with Docling MCP, simply edit the config file claude_desktop_config.json (located at ~/Library/Application Support/Claude/claude_desktop_config.json in MacOS) and add a new item in the mcpServers key with the details of a Docling MCP server. You can find an example of those details here.
Example of prompt for converting PDF documents:
Convert the PDF document at <provide file-path> into DoclingDocument and return its document-key.
Example of prompt for generating new documents:
I want you to write a Docling document. To do this, you will create a document first by invoking `create_new_docling_document`. Next you can add a title (by invoking `add_title_to_docling_document`) and then iteratively add new section-headings and paragraphs. If you want to insert lists (or nested lists), you will first open a list (by invoking `open_list_in_docling_document`), next add the list_items (by invoking `add_listitem_to_list_in_docling_document`). After adding list-items, you must close the list (by invoking `close_list_in_docling_document`). Nested lists can be created in the same way, by opening and closing additional lists.
During the writing process, you can check what has been written already by calling the `export_docling_document_to_markdown` tool, which will return the currently written document. At the end of the writing, you must save the document and return me the filepath of the saved document.
The document should investigate the impact of tokenizers on the quality of LLM's.
Copy the .env.example file to .env in the root of the project.
cp .env.example .envIf you want to use the RAG Milvus functionality edit the new .env file to set both environment variables.
RAG_ENABLED=true
OLLAMA_MODEL=granite3.2:latest
EMBEDDING_MODEL=BAAI/bge-small-en-v1.5
Note:
ollama can be downloaded here
For example we are using granite3.2:latest to perform the RAG search.
To download this model run:
ollama pull granite3.2:latestWhen using the docling-mcp server with RAG this would be a simple example prompt:
Process this file /Users/name/example/mock.pdf
Upload it to the vector store.
Then summarize xyz that is contained within the document.
Known issues
When restarting the MCP client (e.g. Claude desktop) the client sometimes errors due to the .milvus_demo.db.lock file. Delete this before restarting.
The Docling-MCP codebase is under MIT license. For individual model usage, please refer to the model licenses found in the original packages.
Docling and Docling-MCP is hosted as a project in the
IBM ❤️ Open Source AI: The project was started by the AI for knowledge team at IBM Research Zurich.