Summarize Web Pages: Claude + Make

What this does: paste a URL into a Google Sheet and Claude writes a short, ready-to-use newsletter blurb back into the next column.

Time to set up: about 20 minutes once. After that, a blurb appears seconds after you paste a link.

What you need: a Make account (free tier is fine), a Claude API key, and a Google account with a sheet.

Skip if: you write very few blurbs, or the pages you summarize are behind logins that a simple fetch cannot read.

Want a working AI workflow in your inbox each morning? Join the free daily Beginners in AI newsletter.

A person who runs a tiny weekly roundup for their woodworking guild keeps a running list of links worth sharing. The links pile up. Turning each one into a two-sentence blurb for the email is the Sunday-night job that makes them dread the newsletter.

This build does the blurbs. You keep a Google Sheet with a column for links. Paste one, and Make fetches the page, hands it to Claude, and writes a clean blurb back into the next column. By the time you sit down to assemble the email, the summaries are already there.

It is the same toolkit as our other guides, with one extra box: an HTTP step that grabs the page text so Claude has something to read. We build it in Make. If you have done our spreadsheet-to-WordPress build, the spreadsheet trigger will feel familiar.

What does this workflow actually do?

In one line: a pasted link becomes a written blurb. Make watches your sheet. When you add a URL, it fetches the page over HTTP, sends the text to Claude, and writes Claude’s blurb back into the row. No tab-juggling, no copy-paste.

A few real cases, none of them the usual ones:

  • The woodworking-guild roundup above, with Sunday night handed back to the woodworker.
  • A hospital librarian turning new research links into plain-language blurbs for a staff digest.
  • A neighborhood mutual-aid group summarizing resource links into a shareable list for members.
  • A indie bookstore owner blurbing reviewed titles for the shop’s email without reading every full review twice.

Collecting links is easy. Writing the same kind of blurb over and over is the grind. That is exactly the repetitive judgment Claude is good at.

Why use Make if Claude can already write?

Reasonable question. Claude is good at reading a page and writing a tight summary. It is not built to watch a sheet, fetch web pages, and write results back into the right row. That is plumbing, and Make does plumbing.

Make handles the no-judgment steps: spotting the new link, fetching the page, putting the blurb back in place. Claude handles the one step that needs reading and writing, the summary. Make is the conveyor belt; Claude is the one worker who actually reads. I have built versions that asked the AI to do the fetching and bookkeeping too, and they were slower and more fragile. The split keeps it steady.

What do you need before you start?

  • A Make account. The free 1,000 operations a month covers a busy roundup.
  • A Claude API key from the console, not the chat app. See how to use Claude.
  • A Google account with a sheet that has a Link column and a Blurb column.
  • Nothing else. The HTTP step needs no account; it just fetches public pages.

Two Make words. A scenario is the whole four-step automation. A module is one box. This one has four boxes, because fetching the page is its own step.

How does the workflow work, step by step?

Four modules, left to right:

ModuleAppWhat it does
1. TriggerGoogle SheetsFires when you add a row with a link
2. FetchHTTPDownloads the page text at that URL
3. BrainClaudeReads the page and writes a blurb
4. OutputGoogle SheetsWrites the blurb back into the row
Make scenario: Google Sheets to HTTP to Claude to Google Sheets, four connected modules
The finished scenario in Make: a Sheets trigger, an HTTP fetch, a Claude step, and a Sheets write-back, four modules wired left to right.

Step 1: Set up the link sheet

Make a Google Sheet with two columns: Link and Blurb. In Make, add Google Sheets, Watch New Rows, connect your Google account, and point it at the sheet. This trigger fires whenever you paste a new link.

Step 2: Fetch the page

Add an HTTP, Make a request module. Set the method to GET and map the URL to the Link column from Step 1. This downloads the page so Claude has the actual content to summarize, not just the address.

Step 3: Let Claude write the blurb

Add a Claude module, Create a Prompt, paste your key, pick Sonnet, and map the page text from Step 2 into the prompt:

Write a 2-sentence newsletter blurb about this page.
Page text: {{response from Step 2}}

Plain English, no hype. Say what it is and why a reader would click.
Return only the two sentences, nothing else.

Keeping the output to just the two sentences makes the next step clean.

Step 4: Write the blurb back

Add Google Sheets, Update a Row, point it at the same row from Step 1, and map Claude’s blurb into the Blurb column. Now the summary lands right next to the link you pasted.

How do you run it and check the result?

Click Run once. Make reads your test link, fetches the page, summarizes it, and fills in the Blurb column. When I tested this, a blurb appeared a few seconds after pasting a link, and it was usable with at most a word swapped. Tune the Step 3 prompt for length and tone, then turn the scenario on so it runs as you add links through the week.

By newsletter day, your blurbs are written. You go from facing a list of bare links to editing summaries that already exist, which is a different and much smaller job.

What does this cost to run?

PieceFree tierIf you outgrow it
Make1,000 operations/month freeCore plan from about $9/month
Claude APIPay per useA 2-sentence blurb costs well under a cent
Google SheetsFreeFree
HTTPFreeFree

Each link uses four Make operations and a sliver of Claude. For a weekly roundup of twenty links, that is eighty operations a month, well inside the free plan. See our Make guide for tiers.

What can go wrong, and how do you avoid it?

  • The fetch returns junk or nothing. Some pages block simple requests or load content with scripts. Those need a heavier scraping step; most article pages are fine.
  • The blurb is too long or chatty. Tell Claude the sentence count and to return only the blurb, the tweak that cleaned mine up.
  • It writes to the wrong row. Make sure Step 4 updates the row from Step 1, not a fixed cell.
  • Login-only pages. A plain fetch cannot read pages behind a login. Use public URLs.

This four-box pattern adapts to many read-then-write jobs. More in Make AI scenarios.

How do you build this in Zapier or n8n instead?

Same four jobs, different names.

JobMakeZapiern8n
Catch a new linkWatch New RowsNew Spreadsheet Row triggerGoogle Sheets Trigger node
Fetch the pageHTTP requestWebhooks / Code by ZapierHTTP Request node
Write the blurbClaude moduleClaude (Anthropic) actionAnthropic node
Save it backUpdate a RowUpdate Spreadsheet RowGoogle Sheets node

Make and Zapier are friendliest for a first build. n8n handles high volume cheaply. Compare them in Zapier vs Make vs n8n.

Want the ready-made template?

The steps above build it from scratch. If you would rather skip the setup, the importable Make blueprint is yours free: subscribe to the daily newsletter and the download is waiting on the thank-you page, next to our Special Reports. Import it, connect your own accounts, and you are running in minutes.

Subscribe free and grab the template →

Want it set up with you, live?

Book a 1-on-1 Live Claude AI Crash Course and we build your first automation together, screen to screen.

Book the 1-on-1 ($75) →

Want better prompts for it?

The AI Prompt Library has summarizing and blurb prompts ready to paste into Step 3.

Get the Prompt Library ($39) →

A new AI workflow you can copy, every morning

Free daily newsletter. Built for people who want to use AI well, not chase every model.

Free forever. Unsubscribe anytime.

Common questions about this build

Why the extra HTTP step?

Claude needs the page’s text to summarize it, and the trigger only gives it the address. The HTTP step downloads the actual content in between.

Can it summarize PDFs?

A plain fetch struggles with PDFs. For PDFs you add a step that extracts the text first, then send that to Claude.

Can the blurb go to email instead of a sheet?

Yes. Swap Step 4 for a Gmail or newsletter module. Starting with the sheet keeps the build simple and easy to check.

Is the chat app the same as the API?

Same models, different access. The app is for chatting; Make needs an API key.

Will it overwrite my links?

No. It writes only to the Blurb column. Your links stay put.

References and docs

Last reviewed: May 2026. Make, Claude, and Google update their interfaces; check the official pages for exact button names.

You may also like

Discover more from Beginners in AI

Subscribe now to keep reading and get access to the full archive.

Continue reading