Perplexity is turning its AI search engine into something much more personal: a place where your lab reports, step counts, and sleep data can actually talk to each other and answer questions in plain English. With Perplexity Health, the company is betting that the future of health information isn’t just better search results, but better context—your context.

Perplexity Health starts from a simple observation: almost everyone now looks up health questions online, but almost no one has their own medical data in the mix when they do it. Right now, your life is scattered across portals and apps—hospital logins in one tab, lab PDFs in your email, step counts and heart-rate trends buried inside a wearable app. Perplexity’s new product tries to collapse all of that into a single, conversational interface that already feels familiar to anyone who has used the service for general search.

At launch, Perplexity Health plugs into Apple Health, electronic health records from over 1.7 million care providers, and a long list of wearables and health apps, including Fitbit, Ultrahuman, Withings, Clue, and others, via partners like b.well and Terra. That means the same place you ask “Should I be worried about this resting heart rate?” can see your last few months of workouts, your cardiac history in your EHR, and the latest bloodwork your lab pushed into Apple Health. Instead of getting a generic article about “normal heart rate ranges,” you get an answer that is grounded in your actual numbers, wrapped in explanations and citations you can click into.

Under the hood, Perplexity Health is tightly tied to Perplexity Computer, the company’s orchestrator for AI agents that can research, code, and work autonomously in the background. That pairing is what unlocks the more “do things for me” scenarios: build a marathon training plan that accounts for your current fitness level and past injuries, generate a tight visit-prep brief for your next cardiology appointment, or sketch out a personalized nutrition plan that respects both your lab trends and your schedule. In other words, it isn’t just answering questions off the web; it is using agents to continuously pull from a mix of premium medical literature and your personal biomarkers to assemble a tailored response.

The company is also leaning hard into source quality, which is where Perplexity has tried to differentiate itself more broadly in AI search. Health answers draw from clinical guidelines, peer-reviewed journals, and other physician‑grade sources rather than SEO‑optimized “health blogs” designed to rank on Google. Every response carries citations, so if an answer tells you that a certain lab value is mildly elevated, you can drill down into the original guideline or paper rather than taking the AI’s word for it. The product also builds in explicit guardrails—like telling you when a symptom pattern warrants urgent care instead of another round of questions with an AI.

For a company operating in one of the most sensitive data categories on the planet, privacy is the make-or-break detail, and Perplexity is clearly aware of that. Health data is encrypted in transit and at rest, with strict access controls and the ability to disconnect sources or wipe information entirely. Crucially, Perplexity says health data is not used to train its models and is never sold to third parties, a stance that aligns with its broader move away from advertising and toward paid subscriptions and enterprise offerings. In practice, that means Perplexity’s business incentives lean more toward building tools professionals will pay for—like finance and health—than toward squeezing value out of user profiles for ad targeting.

To avoid building this in a vacuum, Perplexity is also standing up a Perplexity Health Advisory Board made up of physicians, researchers, and health tech leaders. Their role is to pressure-test product decisions, assess whether answers clear the bar of evidence‑based medicine, and help design clinical safeguards that feel realistic in a clinic, not just elegant in a product doc. That mirrors a broader pattern in digital health, where serious players are expected to show not just engineers and designers on their team pages, but practicing clinicians who are accountable for how the tools behave in the real world.

There are, of course, limits—Perplexity is explicit that this is an educational tool, not a virtual doctor. It’s not meant to diagnose, treat, or prevent disease, and it repeatedly tells users to talk to a clinician, especially if they are pregnant, nursing, have a history of eating disorders, or live with other medical conditions. But the value proposition is clear: instead of going into an appointment with a vague sense that “something feels off,” you can arrive with a data-driven summary of what changed in your sleep, steps, heart rate variability, or lab markers, plus a list of questions auto‑drafted from reputable sources.

In the near term, Perplexity Health on Computer is rolling out first to Pro and Max users in the United States, with more regions and user groups expected after that. For Perplexity, this is another step in turning its AI search engine into a broader productivity and decision-making platform—one that aims to be good enough that doctors, not just patients, are comfortable having it open on their screens. And for users already used to asking Perplexity everything from “What’s going on with this stock?” to “How do I fix this error in my code?”, the idea of folding in their own health data may feel like a natural, if more intimate, next move.