Product Data Extraction from Website Screenshots using AI & Sheets - n8n Workflow

E-commerce businesses needing automated catalog updates.
Market researchers performing competitive analysis on product pages.
Data analysts looking for automated web scraping and AI-powered extraction tools.
Any user looking for advanced n8n templates integrating AI services and Google Sheets ops.

Manually collecting detailed product information (name, price, rating) from many web pages is time-consuming and prone to errors. This sophisticated n8n workflow solves this problem by completely automating the entire process, starting from a list of URLs in a Google Sheet. It uses the Dumpling AI service to reliably capture high-fidelity screenshots, bypass common web obstacles like cookie banners, and then leverages the AI's vision capabilities to intelligently extract specific, structured product details from that image data. The final step of this n8n automation loop ensures the clean, structured results are immediately saved back into your spreadsheet, providing a scalable and efficient data pipeline. This particular n8n workflow demonstrates the power of combining data triggers with external AI processing.

This entire n8n workflow operates sequentially, beginning with a specific n8n trigger:

1. Monitor Sheet: The n8n trigger waits for new rows to be added to a designated Google Sheet, pulling the product URL (Site column) to initiate the process.


2. Capture Screenshot: The URL is immediately sent to the Dumpling AI API via an HTTP Request n8n node. This API captures the webpage screenshot, automatically handling cookie banners and scrolling to ensure the full page is rendered.


3. Download Image: A subsequent HTTP Request n8n node downloads the generated screenshot file, which is returned in binary format.


4. Base64 Conversion: Since the Dumpling AI extraction API requires the image to be sent as Base64 data, the Convert Image File to Base64 n8n node handles this essential data transformation.


5. AI Extraction: The Base64 image data is sent to the Dumpling AI extraction API along with a detailed prompt instructing the AI to find the product name, price, discount, and star rating. The n8n node is configured to request the output in JSON format for easy parsing.


6. Format Data: The Format Extracted Data for Google Sheets n8n node uses the Set operation to clean and prepare the AI-extracted JSON data, linking it back to the original input item from the n8n trigger.


7. Update Google Sheets: Finally, the Save extracted data to Google Sheets n8n node performs an appendOrUpdate action on the original Google Sheet, adding the newly extracted product details (the AI output) into the spreadsheet.

To use this n8n workflow template, follow these steps:

1. Import the n8n Workflow: Copy the provided JSON code and import it into your n8n instance.


2. Google Sheets Setup: Configure the Google Sheets Trigger n8n node and the final Google Sheets n8n node with the required OAuth2 credentials. Ensure your spreadsheet contains a column for the product URLs (e.g., 'Site') where the n8n trigger will look for new entries.


3. Dumpling AI Credentials: This n8n workflow requires HTTP Header Authentication for Dumpling AI. Obtain an API key from Dumpling AI and configure the httpHeaderAuth credential within n8n.


4. Activation: Once all credentials are set up and connections are tested, activate the n8n workflow.

This n8n workflow relies on several key n8n nodes for its functionality:

Watch for New Screenshot URLs in Google Sheets (googleSheetsTrigger n8n trigger):
Function: Starts the n8n automation when a new row containing a URL is added.
Key Configuration: Polls for rowAdded events. Requires Google Sheets Trigger OAuth2 credentials.

Create Screenshot File via Dumpling AI (httpRequest n8n node):
Function: Calls the Dumpling AI screenshot endpoint, passing the URL dynamically retrieved from the n8n trigger data: {{ $('Watch for New Screenshot URLs in Google Sheets').item.json.Site }}.
Key Configuration: POST method, includes settings like blockCookieBanners: true and autoScroll: true.

Convert Image File to Base64 (extractFromFile n8n node):
Function: Converts the binary image data downloaded in the previous step into a Base64 encoded string, stored in the property data, preparing it for the AI vision model.
Key Configuration: Operation: binaryToPropery.

Extract Text from Screenshot with Dumpling AI (httpRequest n8n node):
Function: Executes the core AI task by sending the Base64 image and a detailed prompt to extract product attributes.
Key Configuration: Prompt: Requests product name, price, discount, and star rating. Specifies jsonMode: true for structured output.

Save extracted data to Google Sheets (googleSheets n8n node):
Function: Appends or updates the Google Sheet with the original URL and the results extracted by the AI n8n node.
* Key Configuration: Operation: appendOrUpdate. Mapping uses expressions to insert data into the Site and product columns.