Perplexity SEO: how Perplexity finds and cites your pages

That matters because every Perplexity answer carries numbered source links, and a page its crawler can't reach can't be one of them. This guide covers what Perplexity says about its index, its two crawlers, why submission and IndexNow don't apply, what you control, how to measure Perplexity visibility, and what nobody can promise.

Where do Perplexity's answers come from?

Perplexity's answers come from a live search of its own web index, which a language model then summarizes with numbered citations. Perplexity's help center says it "uses advanced AI to search the internet in real-time" and that "each answer includes numbered citations linking to the original sources" (Perplexity Help Center).

The index is Perplexity's own. In a September 25, 2025 research article, Perplexity explains that its first answer engine used outside search APIs, and that it later built "our own dedicated search infrastructure" because those APIs were too costly, too stale or too slow (Perplexity Research). The same article describes how the index decides what to crawl, and that is the closest thing to published Perplexity ranking factors:

Perplexity's Search API launch post adds that "each second, our systems process tens of thousands of index update requests" (Perplexity blog). Perplexity hasn't said how it discovers new URLs, whether it reads XML sitemaps, or how it weighs one span against another. Anything more specific than the table above is a guess. For how this compares with ChatGPT, Claude and Google, see how AI search chooses its sources.

Which crawlers does Perplexity use?

Perplexity uses two agents, and only one of them follows robots.txt. Its crawler documentation describes both, and says changes "may take up to 24 hours" to apply (Perplexity docs):

The user-agent strings, as published: Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot) and the same with Perplexity-User/1.0. The research article adds that PerplexityBot "complies with explicit limits set forth in robots.txt files" on request rates.

Allow PerplexityBot without opening everything

A named group replaces the User-agent: * group for that bot. We tested this on October 17, 2026 with Python's standard robots.txt parser against this file:

User-agent: PerplexityBot
Allow: /

User-agent: *
Disallow: /admin/

PerplexityBot    /pricing         allowed
PerplexityBot    /admin/settings  allowed
Perplexity-User  /pricing         allowed
Perplexity-User  /admin/settings  blocked

PerplexityBot could fetch /admin/ because its own group has no Disallow line. Copy your Disallow rules into the PerplexityBot group. Perplexity-User falls under *, but by Perplexity's own description it may fetch anyway, so protect private pages with logins, not robots.txt. The AI crawlers guide has tested rule sets for every major bot, and editing robots.txt on WordPress shows where the file lives on that platform.

Check your firewall, and verify the IPs

Bot protection blocks PerplexityBot more often than robots.txt does. Perplexity's docs give Cloudflare and AWS WAF steps that match on both user agent and IP. In Cloudflare that's Security → WAF → Custom rules, with the action set to Allow (as of October 2026). Perplexity says to fetch its IP lists regularly, because they change. This short script, which we ran on October 17, 2026, checks log IPs against the live lists:

import ipaddress, json, sys, urllib.request

SOURCES = {
    "PerplexityBot": "https://www.perplexity.com/perplexitybot.json",
    "Perplexity-User": "https://www.perplexity.com/perplexity-user.json",
}

def load(url):
    req = urllib.request.Request(url, headers={"User-Agent": "ip-check/1.0"})
    with urllib.request.urlopen(req, timeout=20) as r:
        data = json.load(r)
    return [ipaddress.ip_network(p.get("ipv4Prefix") or p.get("ipv6Prefix"))
            for p in data["prefixes"]]

ranges = {bot: load(url) for bot, url in SOURCES.items()}
for ip in sys.argv[1:]:
    addr = ipaddress.ip_address(ip)
    match = [bot for bot, nets in ranges.items() if any(addr in n for n in nets)]
    print(ip, "->", ", ".join(match) if match else "not a published Perplexity IP")

$ python3 verify_perplexity_ip.py 18.97.1.229 44.208.221.197 203.0.113.7
18.97.1.229 -> PerplexityBot
44.208.221.197 -> Perplexity-User
203.0.113.7 -> not a published Perplexity IP

A request that claims to be PerplexityBot from an unlisted IP isn't Perplexity's declared crawler. Be aware of a public dispute here. On August 4, 2025, Cloudflare said that when its customers blocked Perplexity's declared bots, it saw requests from "a generic browser intended to impersonate Google Chrome on macOS" using IPs outside Perplexity's published ranges (Cloudflare blog). Perplexity replied the same day that "Cloudflare's recent blog post managed to get almost everything wrong" and attributed the traffic to a third-party cloud browser service it "only occasionally uses" (Perplexity blog). We can't settle that dispute. If you want Perplexity to cite you, it rarely matters: allow the declared bots.

Perplexity's crawl is out of your hands, but your Google indexing is a separate question you can answer today: check which of your pages Google has indexed, free.

Can you submit pages to Perplexity?

No. As of October 2026, Perplexity documents no way to submit a URL or sitemap to its index. Its crawler page covers robots.txt, IP lists and firewall rules only, and there is no Perplexity equivalent of Google Search Console or Bing Webmaster Tools.

IndexNow doesn't reach it either. IndexNow (an open "this page changed" notice that Bing and other participating engines accept; Google doesn't use it) shares each submission with its participants. As of October 2026, Perplexity isn't on indexnow.org's list of submission endpoints or its participant file (IndexNow FAQ, searchengines.json). Our IndexNow guide covers what the protocol does reach.

Perplexity's publisher programs are partnerships, not a submission route. The Publishers' Program, announced July 30, 2024, offered launch partners such as TIME and Der Spiegel revenue sharing, free API access and Enterprise Pro (Perplexity blog). Comet Plus, announced August 25, 2025, shares subscription revenue with participating publishers based on "human visits, search citations, and agent actions," and invites publishers to apply at publishers@perplexity.ai (Perplexity blog). Neither post says membership affects which sources Perplexity cites.

What can you control to rank on Perplexity?

You control access, discovery and how quotable your pages are; Perplexity controls the choice. Based on what Perplexity has published:

1. Let PerplexityBot in. Allow it in robots.txt and at your firewall, as above.
2. Make new pages easy to discover. Perplexity doesn't say how it finds URLs, so use the routes every crawler follows: link new pages from pages that already get crawled, such as your homepage, hubs and recent posts, and keep an accurate XML sitemap.
3. Serve the text in HTML. Perplexity describes a parser that extracts "semantically meaningful content"; text that only appears after scripts run, or only inside images, gives it less to work with.
4. Write self-contained passages. Because Perplexity scores "sub-document units" individually (Perplexity blog), each section should open with a sentence that answers its heading on its own. Our AEO guide (answer engine optimization: making pages easy for AI answer engines to find, trust and quote) covers the technique.
5. Keep facts current. Perplexity filters "stale content" and schedules re-indexing by likely update frequency. Update pages when the facts change, not just the date; see content freshness for AI search.
6. Build authority. Perplexity says it prioritizes "documents from authoritative domains." Accurate mentions on sites people already trust are part of that, as our guide to getting recommended by AI assistants explains.

None of this is a Perplexity-only trick. It's the same practice that works for ChatGPT SEO and Claude SEO, applied to an index you can't inspect directly.

How do you measure your visibility in Perplexity?

You measure it yourself, because as of October 2026 Perplexity publishes no visibility or citation report for site owners. Four methods work:

1. Run your prompts in Perplexity. Ask the questions your customers ask, record which sources it cites, and repeat on a schedule, because answers vary between runs.
2. Query the Search API for your domain. Perplexity says its Search API gives "access to the same global-scale infrastructure that powers Perplexity's public answer engine." Its search_domain_filter parameter limits results to up to 20 domains, so a query filtered to your domain shows whether the index returns your pages for that topic. Each result includes date and last_updated fields. As of October 2026 it costs $5.00 per 1,000 requests, and the docs link a playground that needs no API key (Search API docs, pricing). A page the API returns may still not be cited in the app: shared infrastructure doesn't mean identical results, and the model still chooses its sources.
3. Check your server logs. PerplexityBot requests show which pages it crawls; Perplexity-User requests show pages fetched for real questions. Verify the IPs with the script above.
4. Watch referrals and use a tracker. Clicks from Perplexity answers show up as referrals from perplexity.ai in your analytics. Several AI visibility tools track Perplexity citations; our AI visibility tracking guide sets up a weekly routine, and comparison of AI visibility tools lists which tools cover Perplexity, with prices as of October 2026.

What can nobody promise about Perplexity?

Nobody can promise that Perplexity will cite a page, or when. Perplexity decides what it indexes and which spans it shows the model, and it hasn't published weights for any of it.

Timing studies describe averages, not your result. In a July 2026 vendor study of 1,240 sources across 132 brands, measured February–May 2026, MaxAEO, which sells AI visibility monitoring, reported a median of 4 days from publishing a new page on a brand's own site to its first Perplexity citation, the fastest of the seven AI surfaces it tracked (MaxAEO). That's one vendor's medians, not independently replicated, and we have no Perplexity citation data of our own.

What you can control is narrower: let PerplexityBot in, make pages easy to find and quote, keep them current, and measure what Perplexity actually cites. The AI search optimization guide covers the rest of the picture.

Where does FreshlyCited fit?

FreshlyCited works on Google, not on Perplexity. "ChatGPT, Perplexity and other assistants use their own crawlers and search partners. We don't submit to them, and we don't claim to." Being in Google's index doesn't put a page in Perplexity's, which Perplexity builds with its own crawler.

What FreshlyCited does is check whether Google has indexed each of your pages, ask Google to look at the ones it hasn't, and re-check on days 1, 3 and 7. That's the step that makes a page eligible for Google Search, AI Overviews and AI Mode. For sites you verify with an IndexNow key, changes are also sent to Bing. The free plan gives you 10 free links to try, then 100 free checks a month, with no card. Check my pages free.