Back to Templates

Extract Text from Google Drive Images with OCR.space, OpenRouter AI Summary to Sheets & Gmail

Created by

Created by: Tsubasa Shukuwa || tsubasa-shukuwa

Tsubasa Shukuwa

Last update

Last update 2 days ago

Share


How it works

This workflow automatically detects new image files uploaded to a Google Drive folder, extracts Japanese text using OCR, summarizes it with AI, and records the result in Google Sheets.
Finally, it sends a completion email notification with the file name and summary.

Workflow steps:

  1. Google Drive New File Trigger – Watches a specific Google Drive folder for new image uploads.
  2. Download Image File – Downloads the newly uploaded image for processing.
  3. Extract Text with OCR.space – Sends the image to the OCR.space API to extract text (Japanese supported).
  4. Format OCR Result & Check for Empty – Cleans and validates the extracted text.
  5. Generate Summary with OpenRouter AI – Uses an AI model to generate a short summary of the text.
  6. OpenRouter Chat Model – Connects the AI Agent to the OpenRouter language model.
  7. Append row in sheet – Adds the file name, AI summary, and processing date to Google Sheets.
  8. Send Completion Notification via Gmail – Sends an email with the summarized content and Google Sheets link.
  9. Process Completed – Marks the workflow’s successful end.

Setup steps

  1. Connect your Google Drive, Google Sheets, and Gmail accounts through credentials.
  2. Set your OCR.space API key in the HTTP Request node.
  3. Add your OpenRouter API key credential for the AI node.
  4. Replace the Google Sheet ID and folder ID with your own.
  5. Customize the Gmail recipient and email message as needed.
  6. Adjust the polling frequency (e.g., every 1 minute) depending on your workflow needs.

Ideal for

  • Digitizing and summarizing handwritten or printed book pages.
  • Automatically extracting and archiving text from scanned reports or notes.
  • Businesses or educators automating document reading and summarization tasks.

⚙️ Note:
Each node includes a clear English Sticky Note above it for easier understanding and documentation.