Perplexity SEO Tools: Track and Earn Perplexity Citations

Perplexity is the easiest AI engine to measure. Its answers show numbered sources, and its API hands back the same kind of source list as data. It also leans on pages that already rank: in an Ahrefs study of 15,000 prompts, 28.6% of the URLs Perplexity cited sat in Google's top 10, far more than the other assistants tested.

One change makes this week a good time to check your setup. Perplexity's developer docs now say its Sonar chat API "will be supported until September 27, 2026" and recommend the Agent API for all new projects. Any tracking script written for Sonar needs updating.

This guide sorts Perplexity SEO tools by the job each one does, with prices checked on 28 September 2026. For how Perplexity crawls and ranks pages, read our Perplexity SEO guide. For how to read the numbers once you have them, see how to measure GEO.

The Five Jobs Perplexity SEO Tools Do

Most "best Perplexity SEO tools" lists are one long ranking of trackers. That hides the fact that a tracker is only one of five jobs. Here is the whole map, with what each job costs to start.

Use Perplexity SEO tools in that order. A page Perplexity's bots are blocked from reading has no chance of being cited, so access comes first. Tracking comes second because it tells you where to spend effort. The rest is where the results come from.

Job 1: Check That Perplexity's Bots Can Reach You

Access checks are the cheapest Perplexity SEO tools on this list. Perplexity publishes its bots, their user agents and their IP ranges on one crawlers page, and your own logs do the rest.

Two bots, two different signals

• PerplexityBot "is designed to surface and link websites in search results on Perplexity," and Perplexity says it isn't used to crawl content for AI foundation models. It follows robots.txt. Its hits tell you the index can see your pages.
• Perplexity-User fetches a page when a user's question needs it. "Since a user requested the fetch, this fetcher generally ignores robots.txt rules," the same page says. Each hit is a real question that pulled your page, which makes it the closest thing to a citation signal in your own logs.

Perplexity says robots.txt changes can take up to 24 hours to apply. If you block or unblock either bot, wait a day before you judge the result.

Verify hits against the published IP ranges

User agents are easy to fake, so check the source IP of heavy hitters. Perplexity publishes two lists: perplexitybot.json and perplexity-user.json. Each holds a prefixes array of ipv4Prefix ranges. On 28 September 2026 the first listed 8 ranges and the second listed 4. This short script checks an IP against both:

import ipaddress, json, urllib.request
 
LISTS = {
    "PerplexityBot": "https://www.perplexity.com/perplexitybot.json",
    "Perplexity-User": "https://www.perplexity.com/perplexity-user.json",
}
 
def load(url):
    data = json.load(urllib.request.urlopen(url))
    return [ipaddress.ip_network(p["ipv4Prefix"])
            for p in data["prefixes"] if "ipv4Prefix" in p]
 
nets = {bot: load(url) for bot, url in LISTS.items()}
 
def verify(ip, bot):
    addr = ipaddress.ip_address(ip)
    return any(addr in net for net in nets[bot])
 
print(verify("18.97.1.229", "PerplexityBot"))  # True on 28 Sep 2026

Run it on a schedule if your firewall allowlists these ranges. Perplexity's docs tell you to automate the refresh, because the lists change.

Free Perplexity SEO tools for access checks

1. Your logs. Export a day of access logs and paste them into our free AI crawler log analyzer. It counts hits by bot and lists the pages each one fetched, and nothing leaves your browser.
2. A robots.txt test. The robots.txt tester shows whether a rule blocks PerplexityBot on a given URL, and which line did it.
3. A firewall rule. Perplexity's docs give a Cloudflare rule: user agent contains PerplexityBot or Perplexity-User, AND source IP is in the published ranges, then Allow. The docs say a firewall may need this rule before the bots get through.
4. A page scan. The AI search readiness check scans one URL for 12 signals, including crawler access in robots.txt.

Job 2: Trackers That Cover Perplexity, and How They Collect It

Most Perplexity SEO tools on "best of" lists are trackers. A tracker runs your prompts on a schedule and records whether each answer names or links you. For Perplexity, the useful differences sit below the price tag.

Five questions to ask about Perplexity coverage

1. Is Perplexity on the plan you'd buy? Some vendors keep it on a higher tier or sell it as an add-on.
2. Does it query perplexity.ai like a user, or call the API? Perplexity's own API FAQ says results can differ between the two.
3. Does it store the full source list? You want every cited URL per answer, not a yes or no for your brand.
4. How many runs per prompt, and from where? One run is a sample of one, and location changes sources.
5. Can you export raw answers? Raw answers let you check a surprising number yourself.

If a vendor's pages don't answer a question, ask before you buy. To see where trackers sit among other Perplexity SEO tools and AI search tools, read our guide to AI search optimization tools. For the ChatGPT side of the same tools, see our ChatGPT rank tracker guide.

Perplexity SEO tools for tracking, compared

Prices are as listed on each vendor's site on 28 September 2026, except Peec AI's, which come from a Peec page that says they were confirmed on 10 July 2026. "Not stated" means the vendor's public pages don't say.

What each tracker does with Perplexity

Otterly.AI. Lite is $29 a month, or $25 billed yearly, for 15 prompts checked daily. Perplexity is one of four default engines on every plan, with ChatGPT, Google AI Overviews and Microsoft Copilot, and Lite allows unlimited team members. Otterly says it queries as a neutral, non-personalized user. Best for a small team that wants Perplexity from day one.

Semrush AI Visibility Toolkit. It's $99 a month per domain, for 25 custom prompts tracked daily. Perplexity is in the base plan, beside ChatGPT, Google AI and Gemini. Best for teams already running their SEO in Semrush.

Ahrefs Brand Radar. Brand Radar's index holds about 30.9 million Perplexity prompts, so you can look up any brand's Perplexity visibility with no setup. The index starts at $199 a month and is re-tested monthly. For your own prompts, packages start at $50 a month for 2,500 checks, where one check is one prompt, on one platform, in one location, per update. Thirty Perplexity prompts a day for 30 days is 900 checks, well inside that. Best for competitor research at scale.

SE Ranking. The AI Search add-on is $89 a month on monthly billing, on top of a base plan from $129. Its AI Results Tracker lists Perplexity with AI Overviews, AI Mode and ChatGPT, for 200 prompts. Best for agencies that already report from SE Ranking.

Surfer AI Search Analytics. It's $95 a month, or $82 billed yearly, for 50 prompts refreshed daily across ChatGPT, Perplexity, Google AI Mode, AI Overviews and Gemini. Best for content teams who already write in Surfer's editor.

Peec AI. Peec's AI-instructions page says it uses "UI scraping rather than APIs for most tracked engines," so its Perplexity data comes from the interface users see. That page lists Starter at $95 a month for 50 prompts on three models, and says every plan includes Perplexity. Peec's pricing page tells a different story on 28 September 2026. It lists Starter's models as ChatGPT, AI Mode, AI Overviews, Copilot, Gemini and Naver AI, and shows Perplexity under add-on models, which the instructions page prices at $35 a month on Starter. Ask Peec which applies before you buy.

AthenaHQ. The free Essential tier on its plans page gives you 300 credits across ChatGPT, Perplexity, AI Overviews, Gemini and Copilot, with unlimited members. One credit is one AI response, so a 30-prompt panel on all five engines uses 150 credits per run. Starter is $295 a month for 3,600 credits. Best for a free first look.

Profound. Profound says it captures responses "directly from the consumer browsing experience rather than from model APIs." Its free Trial covers ChatGPT, Gemini and Google AI Overviews. Perplexity comes with Enterprise, a custom contract that tracks up to nine engines. Best for large brands with a procurement process.

Job 3: Build a DIY Perplexity Tracker (the Source Ledger)

Paid Perplexity SEO tools save time. A DIY tracker gives you something they may not: every source URL, every search query Perplexity ran, and the cost of each answer, in a file you own. We call this recipe the Source Ledger. It fits a team with a developer and a fixed list of 20 to 50 prompts.

Sonar is now the Agent API

The old way was Sonar: POST https://api.perplexity.ai/v1/sonar with a model such as sonar, returning a citations array of URLs. Perplexity's docs say Sonar "will be supported until September 27, 2026" and recommend "Using the Agent API for all new projects." Its migration guide maps sonar to the fast preset.

Here is what changes if you have an old script:

The auth header stays the same: Authorization: Bearer plus your API key. According to the API reference, each result carries an id, url, title, snippet, date and last_updated. The web search docs say each [n] marker in the answer "refers to a result's id," and the answer's text parts also carry url_citation annotations with the cited URL.

The recipe

Put one prompt per line in prompts.txt, set PERPLEXITY_API_KEY, install requests, and run this once a week. It's written against Perplexity's API reference as of 28 September 2026.

# source_ledger.py: log which sources Perplexity cites for each prompt
import csv, datetime, os, re, time
from urllib.parse import urlparse
import requests
 
API_URL = "https://api.perplexity.ai/v1/agent"
HEADERS = {"Authorization": f"Bearer {os.environ['PERPLEXITY_API_KEY']}"}
DOMAIN = "plannora.io"  # your domain (Plannora is a made-up example)
RUNS = 3
 
def host(url):
    return urlparse(url).hostname or ""
 
def is_mine(url):
    return host(url) == DOMAIN or host(url).endswith("." + DOMAIN)
 
prompts = [p.strip() for p in open("prompts.txt") if p.strip()]
with open("ledger.csv", "a", newline="") as f:
    out = csv.writer(f)
    if f.tell() == 0:
        out.writerow(["date", "prompt", "run", "model", "retrieved", "cited",
                      "my_urls", "cited_domains", "queries", "cost_usd"])
    for prompt in prompts:
        for run in range(1, RUNS + 1):
            r = requests.post(API_URL, headers=HEADERS, timeout=120,
                              json={"preset": "fast", "input": prompt})
            r.raise_for_status()
            data = r.json()
            results, queries, text, cited = [], [], "", set()
            for item in data["output"]:
                if item["type"] == "search_results":
                    results += item["results"]
                    queries += item.get("queries", [])
                elif item["type"] == "message":
                    for part in item["content"]:
                        text += part.get("text", "")
                        cited |= {a["url"] for a in part.get("annotations", [])
                                  if a.get("type") == "url_citation"}
            ids = {int(n) for n in re.findall(r"\[(\d+)\]", text)}
            cited |= {x["url"] for x in results if x["id"] in ids}
            mine = [x for x in results if is_mine(x["url"])]
            out.writerow([
                datetime.date.today().isoformat(), prompt, run, data["model"],
                bool(mine), any(is_mine(u) for u in cited),
                " ".join(x["url"] for x in mine),
                " ".join(sorted({host(u) for u in cited})),
                " | ".join(queries),
                data.get("usage", {}).get("cost", {}).get("total_cost"),
            ])
            time.sleep(1)

What each column tells you

• Retrieved is true when your domain is in the result set Perplexity searched. You were in the running.
• Cited is true when the answer cites one of your pages, either through a url_citation annotation or an [n] marker that points at your result. You won a source slot.
• Cited domains lists every site the answer leaned on. This is your target list for Job 4.
• Queries holds the searches Perplexity ran for the prompt. These are the sub-questions your pages need to answer, a pattern called query fan-out.
• Model matters because presets aren't versioned. Perplexity says calling a preset "always resolves to the latest Perplexity-recommended configuration," so a jump in your numbers may be a model change, not your work.
• Cost comes from usage.cost.total_cost, which Perplexity's pricing page says reports the calculated request cost.

Keep retrieved and cited apart. A page that is often retrieved but rarely cited is close. Rewriting its opening passage may be enough. A page that is never retrieved has an access, indexing or relevance problem first.

What a 30-prompt, 3-run weekly panel costs

On 28 September 2026 the fast preset runs openai/gpt-6-luna on priority processing, per Perplexity's presets page. The changelog says priority "uses 2× the model's standard token prices," and that the preset's search now costs $1.00 per 1,000 calls. The pricing page lists gpt-6-luna at $0.10 per million input tokens and $0.50 per million output tokens. Doubled, that's $0.20 and $1.00.

A 30-prompt panel run three times is 90 calls a week. Token counts vary by prompt, so the table shows two cases. The light case uses the 1,000 input and 500 output tokens that Perplexity's own cost calculator treats as a typical fast run. The heavy case is a round upper bound we chose.

For comparison, the same light call at legacy Sonar's listed rates costs $0.0065: $1 per million tokens each way, plus a $5 per 1,000 request fee. That's about $2.54 a month for the same panel. Either way the API bill is small. The real cost is the hour a week someone spends reading the ledger.

Ninety answers a week is also a small sample. At a citation rate near 10%, the margin of error is at least ±6 points. Pool four weeks into 360 answers and it falls to about ±3 points. Our GEO measurement guide explains the math and a control test for proving your changes worked.

Where the API and perplexity.ai part ways

Perplexity's FAQ says the API "uses the same search system as the UI with differences in configuration," and that "the underlying AI model might differ." Three practical effects follow:

• One search call per answer. The fast preset makes at most one search call and collects up to 10 results. An answer a user sees in the app may be built differently.
• No personal context. Your script has no login, history or location unless you add one. The web_search tool accepts a user_location with a country code if your market is local.
• Filters on by default. The FAQ says SafeSearch is on for the API.

So treat the Source Ledger as a trend line and a source list, next to your other Perplexity SEO tools. For your 5 to 10 most valuable prompts, also check the app by hand now and then. If a question only needs raw ranked results with no answer, Perplexity's Search API returns them for $5 per 1,000 requests, or $1 with Fast Search.

Job 4: Earn the Sources Perplexity Cites

Most Perplexity SEO tools stop at measurement. Tracking and access checks tell you where you stand, but they don't move you. For that you need to be on the pages Perplexity reads, or be one of them.

Turn the ledger into a target list

Sort one week of cited_domains by how many answers each domain appears in. Here is an illustrative week for Plannora, a made-up project management tool, across its 90 answers:

Answers cite several sources, so the counts don't sum to 90. Plannora's own citation rate is 9 in 90, or 10%. That's the baseline its next month is judged against.

Two studies shape how you read a list like this. Profound's analysis of 680 million citations, from August 2024 to June 2025, found that Reddit held 46.7% of the citations going to Perplexity's top 10 domains, but only 6.6% of all its citations. Most sources come from a long tail of ordinary sites. And the Ahrefs overlap above says classic rankings still carry weight here. Your ledger shows which of those forces rules your own prompts.

Tools for earning sources

• Your own pages. Answer-first sections, dated facts and real tables help Perplexity's passage-level ranking pick you. Our guide to optimizing content for AI search covers the writing.
• Community. Reddit and forums matter when the ledger says they do. See Reddit SEO for the rules of engagement.
• Press and reviews. Digital PR and review-site outreach target the third-party domains in your list.
• Branded prompts. If the problem is what Perplexity says about your brand, the fix works differently. Our guide to brand presence in Perplexity traces wrong answers to their sources.

Job 5: Perplexity's Own Programs for Publishers and Stores

Perplexity publishes no citation report for site owners as of September 2026, which is why third-party Perplexity SEO tools exist. It does run programs that pay or feature some sites.

• Comet Plus, for publishers. Announced in August 2025 at $5 a month, and included in Pro and Max, it pays partners on three kinds of use: human visits, search citations and agent actions, per Search Engine Journal. Digiday reported an initial $42.5 million pool with 80% going to publishers. Publishers apply by emailing publishers@perplexity.ai. Check current terms with Perplexity before you plan around them.
• The Merchant Program, for stores. It launched in November 2024 with Perplexity's shopping feature. Merchants who share product details give Perplexity more complete data to recommend from, and at launch they got free API access for search on their own sites.
• Instant Buy with PayPal. Since 25 November 2025, US users can buy inside Perplexity, and PayPal's Store Sync makes merchant catalogs discoverable there.

Like the other Perplexity SEO tools here, none of these buys a place in an answer. They add revenue or product data on top of the organic work.

Which Perplexity SEO Tools Fit Your Team

This decision table matches Perplexity SEO tools to teams. Find the row that looks like you, start with the first column, and add the second only when the first stops being enough. Prices are as listed on 28 September 2026.

Whichever Perplexity SEO tools you pick, keep the Source Ledger idea: log sources, not only mentions. The source list tells you what to do next.

Where Rankbox Fits in a Perplexity Stack

Rankbox isn't one of the Perplexity SEO tools for tracking, and it doesn't track AI citations today. It belongs in Job 4. Answer-Space Research maps the questions buyers ask ChatGPT, Perplexity and Google, with model-estimated volume, difficulty and intent. The Citation-Ready Writer researches the live web and writes 2,000 to 3,500-word articles with their sources cited, and every draft gets an SEO and GEO score. Articles reach your site through Rankbox's API, which a developer wires in once. Reddit Presence, which finds threads and drafts replies for you to post, is rolling out.

The Business plan is $49.50 a month for one site and 30 articles, with a 7-day trial of up to 7 articles (a card is required). Pair it with any tracker above, or with the Source Ledger, and aim each article at a prompt where your ledger says you're absent. See plans and pricing.