Perplexity launches web-monitoring API that alerts apps to relevant changes
Developers can watch broad topics or specific pages, with adjustable relevance checks deciding which updates reach their applications.
Loading page…
Developers can watch broad topics or specific pages, with adjustable relevance checks deciding which updates reach their applications.
Listen to this story
Perplexity’s Monitor API, launched Nov. 16, 2026, lets developers set recurring web or page monitors and route qualifying changes into apps and agent workflows through webhooks. Topic monitors search web and news results; page monitors track up to 10 URLs against a saved baseline and report text diffs. Pricing is $1 per 1,000 successful topic-monitor runs or fetched page URLs. Perplexity reports favorable results in an eight-hour, 60-topic comparison, but the assessment was company-run and used a separate model, not an independent test.
The Decider system filters web results for relevance, source legitimacy, new information and freshness; developers can adjust thresholds and change a monitor’s goal without recreating it.
Checks can run every 15 minutes, hourly, daily, weekly or on a custom cron schedule, with duplicate results removed across runs.
Signed webhooks retry after temporary delivery failures, and monitors can alert developers when automatically paused.
Applications can now ask Perplexity to watch for web changes and send relevant updates, rather than repeatedly request searches themselves. Launched November 16, 2026, the Monitor API checks broad topics or specified pages on a schedule, then delivers alerts through webhooks—automated messages that can start work in another application or AI agent.
Perplexity is pitching the service as an alternative to scheduled searches supported by custom code for removing duplicates and spotting changes. Its examples include a coding agent watching for breaking API changes and a product team following a competitor’s pricing page.
Developers define a monitor with a search query or a list of pages, plus an optional goal written in plain language. The two modes differ in what they watch and what they return:
Finding something new does not automatically generate an alert. Perplexity says its Decider system checks web results against four criteria: relevance, source legitimacy, new information and freshness. An update goes out only when every check clears its threshold. For page monitoring, Decider instead checks whether the change is meaningful for the monitor’s goal.
Developers can adjust each check’s threshold, and every update includes a confidence score. They can also rewrite the goal without recreating the monitor. That gives an application a way to change what counts as relevant while keeping the existing monitor in place.
The proposed uses extend beyond notifications. Perplexity describes competitor updates feeding competitive briefs, and regulatory announcements being summarized and routed to compliance teams for review. A webhook can also trigger an agent workflow, such as recommending pricing changes after a competitor update.
Web monitoring is billed by successful run. A run is one scheduled or on-demand check.
Page monitoring is billed by URLs fetched, because one run can fetch multiple pages.
Monitors can run every 15 minutes, hourly, daily or weekly, or follow a custom cron schedule—a rule specifying when checks should run. Each run picks up where the previous one ended, and the service removes duplicate results across runs. Developers can pause, resume or run monitors on demand through the API or API Console.
Relevant updates arrive through signed webhooks, with automatic retries for temporary delivery failures. Monitors can also notify developers when they are automatically paused. Metadata attached to updates can route them back to the appropriate user, workspace or agent.
Perplexity also claims an advantage over Search API providers in its own evaluation. It monitored 60 topics drawn from production search queries for eight hours, then used a separate model to assess relevance and publication timing. The company says Monitor returned the most fresh, relevant results across the widest range of topics. That is a company-run comparison, with another model judging those qualities, rather than an independent test.
The Monitor API is available now through the API and API Console. Existing developers can use the same API key they already use for Perplexity’s Search, Context and Agent APIs.
Loading discussion...
Join the conversation
Explain when fewer alerts would be worth the risk of missing something.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.