Skip to main content
In this quickstart, you import an n8n workflow that downloads files over HTTP and uploads them to Pinecone Assistant. You then chat with the files through an n8n AI agent, which uses an OpenAI model to answer questions with context retrieved from your assistant. To follow along, you need a Pinecone account, an n8n account, and an OpenAI API key.
1

Create an assistant

In the Pinecone console, create an assistant named n8n-assistant. The workflow template looks for an assistant with this name.
2

Install the Pinecone Assistant node

In your n8n account, open the nodes panel, search for Pinecone Assistant, and install the node.
If the Pinecone Assistant node doesn’t appear in the nodes panel, restart your n8n workspace.
3

Import the workflow

Copy the workflow template URL:
In your n8n account, create a new workflow and paste the URL anywhere in the workflow editor. Then click Import to add the workflow.
4

Connect the nodes to your accounts

The workflow has two Pinecone Assistant nodes, Upload file to Assistant and Get context from Assistant. In each one, do the following:
  1. Click Connect to Pinecone and sign in to your Pinecone account. If the button doesn’t appear, as on self-hosted n8n, create a Pinecone credential with your Pinecone API key instead.
  2. For Assistant Name, select n8n-assistant.
Then open the OpenAI Chat Model node, select Credential to connect with > Create new credential, and paste in your OpenAI API key.
5

Upload files

Click Execute workflow. By default, the workflow downloads Pinecone’s release notes and uploads them to your assistant.
To upload your own files, change the URLs in the Set file urls node.
6

Chat with your files

After the files finish uploading, click Open chat at the bottom of the workflow editor and ask a question about them:
The AI agent retrieves relevant context from your assistant and uses it to answer.
7

Clean up

When you no longer need n8n-assistant, delete it in the Pinecone console. Deleting an assistant also deletes all files uploaded to it. You can also delete the imported workflow in n8n, so it doesn’t run by accident later.

​
Next steps

You can customize the workflow for your own use case:
  • Change the URLs in the Set file urls node to use your own files.
  • Edit the system message in the AI Agent node to describe what kind of knowledge your assistant holds.
  • Add the Top K or Snippet Size parameters to the Get context from Assistant node to manage token usage.
  • Add a Metadata Filter to the Get context from Assistant node to narrow the context it retrieves.
To go further: