Have a browser agent locate and tap add to cart

Víctor Paytuvi used Claude computer use to click a shopping site’s add-to-cart control after Gemini had only searched for it.

Done with Claude computer useGemini· Beginner· Oct 2025

What they did
Víctor asked computer-use models to inspect a website for him. Gemini was tried first and spent a long time scanning before it found the purchase button. Claude computer use then handled the same page and both found and clicked the button without needing more direction.
How it went
Claude finished on the first attempt. Gemini never reached the click, and Víctor judged Claude to be more capable and more self-directed.
Tips
Give the agent the page, the product to look for, and any choices it needs to make before purchase. Tell it to stop and wait before any checkout step or other irreversible action.

Try it yourself with Claude computer use

Open [website URL] in a browser. Locate [product name or description], select [size/color/options] if those choices are present, then press the add-to-cart button. Once the item has been added, pause and report the product name, price, and quantity visible on the page. Do not checkout, enter payment information, submit an order, delete anything, or send anything unless I first say to continue.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on x.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.