Analyze PDFs & Create Excel Summaries with a Local LLM on Mac

8/12/2025

The Ultimate Guide: How to Analyze PDFs & Create Excel Summaries with a Local LLM on Your macOS

Ever found yourself drowning in a sea of PDFs? Whether you're a student wading through research papers, an analyst dissecting financial reports, or just someone trying to make sense of a pile of documents, the struggle is real. What if you could have your own private AI assistant on your Mac to read all those PDFs & spit out a neat Excel summary? Sounds pretty cool, right?

Well, it's not science fiction. It's totally doable, & I'm going to walk you through how to set it up. We're talking about running a large language model (LLM) LOCALLY on your machine. This means your data stays with you, no subscriptions, no API keys, & no sending sensitive information over the internet. It's 100% free & private.

We'll be using a popular tool called Ollama to run the LLM, a bit of Python to glue everything together, & you'll be batch-summarizing hundreds of PDFs in no time. Let's dive in.

Why Bother with a Local LLM?

Before we get our hands dirty, let's talk about why this is such a game-changer. Running an LLM on your own Mac offers some serious advantages, especially when you're dealing with sensitive or confidential documents.

Privacy is EVERYTHING: Your data never leaves your computer. This is a huge deal for anyone working with internal reports, client information, or any other data you wouldn't want to upload to a third-party service.
Total Control: You have full command over the model & the data. You can tweak, customize, & optimize the process to fit your exact needs.
No Costs or API Keys: Forget monthly subscriptions or paying per-token. Once you have the setup, it's completely free to run as much as you want.
Offline Capability: Your AI assistant works even without an internet connection.

Honestly, the ability to create a powerful, private data analysis tool on your own hardware is one of the most exciting developments in AI right now.

The Tech Stack: What You'll Need

Here's a quick rundown of the tools we'll be using. Don't worry if some of these are new to you; I'll walk you through the setup.

Ollama: This is the magic ingredient. Ollama is a tool that makes it incredibly easy to download & run powerful open-source LLMs right on your macOS. You can think of it as a local server for your AI models.
A Language Model (like Google's Gemma 3): We need an actual brain for our operation. Gemma is a family of lightweight, state-of-the-art open models from Google. We'll be using a smaller version that's perfect for running on a personal computer.
Python: The glue that holds this all together. We'll use a Python script to automate the process of reading PDFs, sending the text to our local LLM, & creating the Excel summary.
Essential Python Libraries:
- 1pandas
  : For creating & managing our final Excel spreadsheet.
- 1PyPDF2
  : A library for reading & extracting text from PDF files.
- 1olama
  : The official Python library to communicate with our running Ollama instance.
- 1sentence-transformers
  : This helps us convert text into numerical representations (embeddings) which is part of a more advanced technique called Retrieval-Augmented Generation (RAG) for better accuracy.
- 1numpy
  : A fundamental package for scientific computing in Python, used here for vector math.
- 1openpyxl
  : A library required by pandas to handle the .xlsx Excel file format.

Step 1: Setting Up Your Local AI Powerhouse with Ollama

First things first, let's get Ollama & a language model running on your Mac.

Download & Install Ollama: Head over to the Ollama website & download the macOS version. The installation is straightforward, just like any other Mac app.
Open Your Terminal: The Terminal is your command center for this part. You can find it in
1Applications/Utilities/Terminal
.
Pull the LLM: Once Ollama is installed, you need to download a model. We'll use a smaller, efficient version of Google's Gemma 3. Type the following command into your Terminal & hit Enter: