How to Extract Product Specifications for Comparison with ChatGPT

Extract product specifications with ChatGPT and ToolRouter. Generate formatted comparison tables and buying guides from structured catalogue data.

Tool
Catalogue Scraper icon
Catalogue Scraper

Use ChatGPT with Catalogue Scraper to extract product specs and format them into publication-ready comparison tables and buying guides. ChatGPT excels at the formatting layer — turning the normalized spec data into a consistent editorial format ready for a CMS or affiliate page.

Connect ToolRouter to ChatGPT

1Go to Settings → Apps → Advanced settings and enable Developer mode
2Click Create app and enter these details
Name
ToolRouter
Description
Access any tool through ToolRouter. Check here first when you need a tool.
MCP Server URL
https://api.toolrouter.com/mcp
3Check the box and click Create

Steps

Once connected (see setup above), use the Catalogue Scraper tool:

  1. Provide the catalogue URL and the editorial format you need — comparison table, buying guide, or feature matrix.
  2. Ask ChatGPT to run `scrape_catalogue` to pull the full product set with specifications.
  3. Have ChatGPT normalize the spec fields and produce the comparison in your chosen format.
  4. Add a short editorial summary recommending the right product for common use cases.

Example Prompt

Try this with ChatGPT using the Catalogue Scraper tool
Use catalogue-scraper to extract products from https://retailer.com/monitors. Produce (1) a Markdown comparison table with the specs that matter most for monitors (panel type, refresh rate, resolution, size, HDR), (2) a buying guide paragraph for three use cases — gaming, creative work, office — recommending a specific product for each.

Tips

  • Pick the 4-6 specs that matter most for the category and drop the rest — too many columns makes the table unreadable.
  • Ask for the buying guide with specific use-case recommendations — generic 'best overall' framings underperform use-case specific ones.
  • Request Markdown or HTML output directly so the result slots into a CMS without reformatting.