0.25¢ per URL

Let Claude Coderead the page.

Point it at a public URL and get the title, description, social image, canonical and favicon back. A quarter of a cent a call, with no scraper of your own to keep working.

TaskFuel connects Claude Code to paid tools through one prepaid balance. You review the provider and request quote before spending.

Start for FreeWhat it can do ↓
No card required
Works with

What your agent can call.

Currently available through 2s. Your AI picks the one that fits and checks the exact price before it pays.

  • Extract URL metadata

    Extract a public page’s metadata, including its title, canonical URL and social image. Only fields present in the fetched page can be returned.

    0.25¢

Try it in your next task.

Sign in to TaskFuel and copy the connection prompt into your agent. Approve the connection, then ask for the work you need.

Connect TaskFuel in the environment where Claude Code is working. Keep the request with the project context and tell Claude Code where to save the result.

01

Name the task

Tell your agent what the output needs to contain.

02

See the price

Ask for the provider and request quote before spending.

03

Review the response

Check the output against your task. The call reports its cost and remaining balance.

When to use it.

Before you cite it

Check the link is what you think.

A URL in a commit message or a doc can be stale or wrong. One call at 0.25¢ says what the page actually is.

Link previews

Build the unfurl, do not write the parser.

Every link-preview feature starts by reimplementing Open Graph parsing badly. This returns the structured fields directly.

Checking your own routes

Find the missing og:image.

Run every route through it and see which pages unfurl as a blank rectangle. At 0.25¢, checking a hundred costs a quarter.

Providers for this task.

CapabilityProvidersReference
Preview a link2sCompare routes · Endpoint data

Good Questions.

What does Claude Code get back?

The page title, description, social image, canonical URL, favicon, author and other structured metadata, as one quoted call.

What does it cost?

0.25¢ per URL. Checking a hundred routes costs about a quarter.

Why not just fetch the page?

You can, and then you own the parser, the redirects and the sites that need JavaScript. This returns the fields already extracted.

Do I need a scraper account?

No. One TaskFuel account and one prepaid balance covers it.

What if a URL fails?

Upstream 4xx and 5xx responses are not charged. A successful provider response can still be charged when its answer is empty or unhelpful.

Your agent can do things it couldn’t this morning.

$1 in free credit, no card required.

Start for Free