🤖 Visual Regression Testing with Apify & AI Vision

Description

Automate screenshot comparisons of your web pages using AI-powered image analysis. This workflow performs visual regression testing by comparing past and present webpage screenshots using multimodal LLMs inside n8n.

Perfect for frontend QA teams, devops, or product managers who want to catch visual bugs without manual review.

🧠 How It Works:

✅ A Google Sheet contains the list of webpages to monitor ✅ 📸 Phase 1: Apify captures baseline screenshots of each page ✅ Screenshots are uploaded to Google Drive as reference images ✅ 🕒 Phase 2: A scheduled trigger runs at a future time (daily/weekly) ✅ Fresh screenshots are taken using Apify again ✅ Old (baseline) and new screenshots are retrieved from Drive ✅ Both images are passed to an LLM node using binary inputs ✅ The LLM compares the two screenshots based on provided visual criteria ✅ AI outputs detected visual regressions or confirms visual match ✅ Final comparison reports are generated and optionally uploaded to Linear or shared via Slack

🔍 It Automates:

✅ Website visual regression testing across time ✅ Comparison of webpage versions with AI vision ✅ Identification of UI bugs, styling issues, or broken layouts ✅ Optional: Report uploads to task managers (Linear, Jira, etc.)

💡 Why Choose This Workflow:

✅ Uses AI to go beyond pixel-diff—understands layout, structure, and UI changes ✅ No need for complex testing frameworks or manual comparison ✅ Works well for marketing pages, product UI, or web dashboards ✅ Highly customizable and low-maintenance for ongoing QA

👤 Who Is This For:

✅ Frontend and QA engineers ✅ DevOps and SRE teams doing release verification ✅ Product managers tracking visual consistency ✅ Design teams monitoring brand alignment ✅ Agencies managing multiple client sites

🔗 Integrations:

✅ Apify.com (or any screenshotting service) ✅ Google Sheets (list of target URLs) ✅ Google Drive (image storage) ✅ Multimodal LLM node in n8n (e.g., GPT-4 Vision, Claude, etc.) ✅ Optional: Slack or Linear for report delivery

Import this into n8n, connect your Apify and Google accounts, and let your AI-powered visual regression workflow start validating your UI—pixel-perfect (and beyond).

Link : [https://lovable.dev/projects/598407f7-c460-4545-bda0-edbd3a4871e5]