Introduction to PageAssist
PageAssist is an open-source Chrome extension that provides a simple interface for using Ollama directly in the browser. It also supports features such as RAG (Retrieval-Augmented Generation) and MCP (Model Context Protocol).
In this article, I will introduce how to implement RAG using PageAssist and Ollama. I will use Lanner Electronics’ first-quarter financial report for 2026 as an example, allowing the language model to answer questions based on the content of the uploaded document.
Download PageAssist
First, open the Chrome Extensions page and click “Visit Chrome Web Store.” Search for PageAssist and download the extension.
As shown in the figure below, this is the download page for PageAssist.
Open PageAssist
After downloading and installing PageAssist, click the button shown below to open the PageAssist window.
Once opened, you will see that PageAssist provides an interface similar to Ollama.
In the area highlighted by the red box in the figure below, you can select the Large Language Model (LLM) you want to run. In this example, I use the Gemma 4 model.
Next, click “Settings” in the upper-right corner.
Configure PageAssist
PageAssist uses English as the default interface language. If you prefer to use Chinese, you can select Traditional Chinese from the Language option under General Settings.
You can also select your preferred search engine in the web search settings.
Configure RAG and the Embedding Model
Next, click Pipeline in the settings. On the left side, you will see the RAG-related settings.
RAG requires an Embedding Model to convert document content into vectors, which can then be used for document retrieval. In this example, I use:
nomic-embed-text-v2-more:latest
After selecting a suitable Embedding Model, you can proceed to create the knowledge base required for RAG.
Create Knowledge
Click “Knowledge Management” in the left-hand menu to select the documents you want to add to your Knowledge base.
Next, click “Add Knowledge” on the right side. A file upload window will appear.
Select the document you want to add. PageAssist supports various file formats, including PDF, CSV, TXT, and DOCX.
Upload the Financial Report
In this example, I use Lanner Electronics’ first-quarter financial report for 2026 as the data source for RAG.
After uploading the document, the status will initially be displayed as “Processing.” You need to wait until the status changes to “Completed”, which indicates that the document has been successfully processed and is ready for queries.
Query the Document Using RAG
After the document has been uploaded and processed, select the document you just created and uploaded in the Prompt window.
You can then ask questions about Lanner Electronics’ first-quarter financial report for 2026.
For example, you can ask about the operating revenue information provided in the financial report.
PageAssist retrieves relevant information from the uploaded PDF document and provides it to the language model. The model then generates an answer based on the retrieved information.
As shown in the figure below, the model can answer questions about Lanner Electronics’ first-quarter operating revenue based on the uploaded PDF document.
Conclusion
PageAssist provides a relatively simple way to build RAG applications with Ollama.
In this article, I used Lanner Electronics’ first-quarter financial report for 2026 as an example. I added the PDF document to the Knowledge base, used an Embedding Model to enable document retrieval, and finally allowed the LLM to answer questions based on the content of the document.
If you would like to learn more about how RAG works behind the scenes, I recommend first reading an introduction to RAG and Embeddings and then trying PageAssist yourself. This makes it easier to understand the overall workflow.