How can I automate Pinterest pin creation using Google Sheets data and n8n?

Below is an example of how you can build this Pinterest automation workflow in n8n, even though there isn’t a dedicated Pinterest node. You’ll need to leverage n8n’s core nodes (like Google Sheets and HTTP Request) to interact with external APIs such as Pinterest and Hugging Face.

---

1. Get Data from Google Sheets

• Use the Google Sheets node to read your spreadsheet.
• Configure it to pull the columns (pin title, pin status, description, destination URL, board name) you need.
• You may want to filter the rows by status (for example, those marked as “pending”) using a Function or IF node.

---

2. Search for the Pinterest Board

• Since there isn’t a built-in Pinterest node, use an HTTP Request node to call the Pinterest API endpoint that searches for a board.
• Pass the board name from the Google Sheet as a query parameter (or in the request body) according to Pinterest’s API documentation.
• Ensure you have set up any needed API credentials (perhaps OAuth tokens or API keys) in the node’s authentication parameters.

Example configuration for the HTTP Request node:

JSONCopied
{
  "method": "GET",
  "url": "https://api.pinterest.com/v3/boards/search",
  "qs": {
      "query": "={{ $json.boardName }}"
  },
  "headers": {
      "Authorization": "Bearer YOUR_PINTEREST_API_TOKEN"
  }
}

---

3. Create an Image using a Hugging Face Request

• Use another HTTP Request node to call a Hugging Face inference endpoint to generate your image.
• Configure the node to send any required input (such as a prompt or additional parameters).

Example configuration:

JSONCopied
{
  "method": "POST",
  "url": "https://api-inference.huggingface.co/models/YOUR_MODEL",
  "headers": {
      "Authorization": "Bearer YOUR_HF_API_KEY",
      "Content-Type": "application/json"
  },
  "body": {
      "inputs": "Generate an image for {{ $json.pinTitle }} with {{ $json.description }}"
  },
  "json": true
}

The response should include your generated image (or a URL/link to it) which you can use in the next step.

---

4. Post the Pin to Pinterest

• Use a third HTTP Request node to post the pin.
• In the request body, send over your pin title, description, destination URL, generated image, and the board ID (from step 2).
• Again, configure the proper authorization headers for Pinterest.

Example configuration:

JSONCopied
{
  "method": "POST",
  "url": "https://api.pinterest.com/v1/pins",
  "headers": {
      "Authorization": "Bearer YOUR_PINTEREST_API_TOKEN",
      "Content-Type": "application/json"
  },
  "body": {
      "board": "={{ $json.boardId }}",
      "note": "={{ $json.pinTitle }} - {{ $json.description }}",
      "link": "={{ $json.destinationUrl }}",
      "image_url": "={{ $node[\"Hugging Face Request\"].json.image_url }}"
  },
  "json": true
}

(Note: Replace the URL and body parameters according to the latest Pinterest API documentation.)

---

5. Update Google Sheets to Mark the Task as Finished

• Finally, use another Google Sheets node to update the row.
• Change the status from “pending” to “finished” (or another status) so you have record of the completed pin.

Example update using the Google Sheets node:

JSONCopied
{
  "sheetId": "YOUR_SPREADSHEET_ID",
  "range": "Sheet1!A2:F2",  
  "values": [
    ["New pin title", "finished", "Updated description", "new destination url", "board name", "other info"]
  ]
}

(Adjust the range and values as needed based on your sheet’s structure.)

---

Final Workflow Overview

1. Google Sheets (Read data) →
2. Filter/Split the Data (if needed) →
3. HTTP Request (Search for Pinterest board) →
4. HTTP Request (Generate image with Hugging Face) →
5. HTTP Request (Post pin to Pinterest) →
6. Google Sheets (Update pin status)

With this multi-step flow in n8n, you can automate your entire process even when a dedicated Pinterest module isn’t available.

Be sure to consult the respective APIs’ documentation (Pinterest and Hugging Face) for any details on endpoints, request formats, and authentication methods.

If you have any questions on setting up any of these individual nodes or further details on n8n workflows, feel free to ask!