Batch Image Upscaling using AI and Google Drive - n8n Workflow

• Photographers and Editors: Seeking a scalable solution to batch upscale portrait photos without manual intervention.


• Data Managers: Users who store image assets in Airtable and need automated processing pipelines.


• n8n Automation Developers: Looking for advanced n8n templates that combine cloud storage, database access, and external AI APIs.


• Businesses needing a highly efficient, high-volume image enhancement n8n solution.

Managing large volumes of media assets, especially when needing enhancement like upscaling, can be time-consuming. This specialized n8n workflow solves this by creating a robust pipeline that links Airtable, a leading AI image processor (Real-ESRGAN via Replicate), and Google Drive.

Upon manual execution of this n8n trigger, the system dynamically creates a dedicated output folder in Google Drive. It then pulls specified image URLs from an Airtable record. Each image is processed individually through a looping mechanism, sent to Replicate for high-quality upscaling, and the final high-resolution file is downloaded and immediately uploaded back to the newly created Google Drive folder. This provides a clean, fully automated, and highly repeatable AI enhancement process powered by n8n.

This comprehensive n8n workflow operates through eight main stages:

1. Start: The process begins with the manual n8n trigger, executed by clicking ‘Execute workflow’.


2. Folder Creation: The Google Drive n8n node creates a new, uniquely named folder to store the upscaled images.


3. Data Retrieval: The Airtable n8n node fetches a specific record containing the source images (expected in the PortraitFotoAuswahl field).


4. URL Extraction: A specialized Code n8n node processes the Airtable output, extracting individual image URLs and attaching the Google Drive folder ID created in Step 2 to each item.


5. Batch Looping: The Split In Batches n8n node iterates over every extracted URL, ensuring each image is processed sequentially.


6. AI Upscaling: The HTTP Request n8n node sends the image URL to the Replicate API, leveraging the Real-ESRGAN model for 2x upscaling.


7. File Download: A subsequent HTTP Request n8n node downloads the resulting upscaled image file provided by the Replicate service.


8. Final Upload: A second Code n8n node ensures the necessary metadata (specifically the target folder ID) is correctly attached to the binary file. Finally, the Google Drive n8n node uploads the finished upscaled image to the designated folder. This entire process defines an efficient image processing n8n workflow.

To deploy this n8n template, follow these steps:

1. Import: Copy the provided JSON data and import it directly into your n8n instance.


2. Google Drive Credentials: Configure the Google Drive n8n node connections using your existing OAuth2 credentials. Ensure the account has permission to create folders and upload files.


3. Airtable Setup: Connect the Airtable n8n node by providing your API token. You must specify the Base ID, Table ID, and the exact Record ID containing the portrait images. The images must be stored in an attachment field named PortraitFotoAuswahl.


4. Replicate API Key: Set up an HTTP Header Auth credential named 'Replicate' for the 'Replicate Upscaler' n8n node. This typically involves setting the Authorization header with your Replicate API key (e.g., Token APIKEY> or similar configuration required by Replicate).


5. Configuration: Update the Folder Name in the 'Create folder' n8n node and verify the Record ID in the 'Get Record from Pictures' n8n node before executing the n8n workflow.

When clicking ‘Execute workflow’ (Manual Trigger): The starting point, acting as the primary n8n trigger for initiating the batch process.
Create folder (Google Drive n8n node): Dynamically creates a folder in Google Drive using a user-defined name template, ensuring organization for the output files.
Get Record from Pictures (Airtable n8n node): Fetches the specific Airtable record data needed to extract image attachments. Requires careful configuration of the Base, Table, and Record IDs.
Extract Portrait URLS from Airtable Output (Code n8n node): Critical custom script that parses the PortraitFotoAuswahl field, extracts each image URL, and maps the newly generated Google Drive folder ID for later use in the n8n workflow.
Loop Over Items (Split In Batches n8n node): Manages the batch processing, ensuring the subsequent AI steps are performed individually for every extracted image URL.
Replicate Upscaler (HTTP Request n8n node): The core AI processing n8n node. Sends a POST request to Replicate using the Real-ESRGAN model version specified in the JSON body, aiming for 2x scaling. This n8n node handles the demanding upscaling task.
Download Upscale (HTTP Request n8n node): Downloads the resulting upscaled image file from the temporary URL provided by Replicate, preparing the binary file for storage.
Upload to Google Drive (Google Drive n8n node): The final action. Uploads the processed binary file to the correct, dynamically referenced folder ID using a timestamped filename, completing the n8n workflow.