Firehose
Get started

Stream vs URL Watch

Firehose monitors the web two ways — Stream matches the whole crawl against your rules, URL Watch tracks a fixed list of pages. Pick the right one.

View as Markdown

Firehose gives you two ways to monitor the web, and they answer different questions:

  • Stream — "Tell me about any page on the web that matches this pattern."
  • URL Watch — "Tell me when these specific pages I already know about change."

The difference comes down to one thing: who decides which pages get looked at.

How each one sees pages

A Stream tap sits on top of Firehose's continuous web crawl. It evaluates pages as the crawler reaches them, on the crawler's own schedule — the crawler decides when any given page is visited. That makes it ideal for discovery across the whole web, and the wrong tool for watching one page you care about, because nothing makes the crawler fetch that page on demand.

URL Watch works the other way around. You hand it a specific list of URLs, and it re-crawls each one on a cadence you choose, recording the diff between crawls. The pages looked at are exactly the ones you named, on a schedule you set.

At a glance

StreamURL Watch
What you monitorThe whole web, by patternA fixed list of URLs you choose
What you defineRules (queries) on a tapURLs + a crawl frequency
What triggers a checkThe crawler reaching a pageYour chosen cadence
How you receive resultsLive SSE stream (and the dashboard feed)Crawl history with diffs in the dashboard
BillingPer matchPer check, against a monthly quota

A common pitfall

Pointing a Stream rule at one URL (url:"https://example.com/pricing") does not make Firehose fetch that page — it only narrows which crawled pages match. If the crawler hasn't re-crawled that page, a change to it never reaches your stream. To watch a known page for changes, that's exactly what URL Watch is for.

Next steps