Perplexity Scraper - Threads, Answers, Sources & Discover avatar

Perplexity Scraper - Threads, Answers, Sources & Discover

Pricing

$4.50 / 1,000 results

Go to Apify Store
Perplexity Scraper - Threads, Answers, Sources & Discover

Perplexity Scraper - Threads, Answers, Sources & Discover

Scrape public Perplexity threads and Discover stories: every answer, its cited sources (title, URL, snippet, date), related questions, the searches Perplexity ran, model and mode, plus Discover articles with full text and sources. Brand-mention tracking and monitor mode. No login, no browser.

Pricing

$4.50 / 1,000 results

Rating

0.0

(0)

Developer

Scrapers Delight

Scrapers Delight

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

a month ago

Last modified

Categories

Features

Limited permissions

Share

What does Perplexity Scraper do?

Perplexity Scraper turns public Perplexity answers into clean rows. Give it the link of any public (shared) Perplexity thread and you get every answer in it: the question, the full answer, every cited source with its title, URL, snippet and date, the related questions Perplexity suggested, the web searches it ran to write the answer, the model and the mode. Pick Perplexity Discover topics and you get each Discover story with its summary, key points, full article text and every source behind it.

Also known as: Perplexity thread scraper, Perplexity Discover scraper, Perplexity AI answers scraper, Perplexity Pages scraper.

  • Every answer in a thread - the first question and every follow-up, one row each, with the thread's views and forks.
  • Sources the way Perplexity cites them - sources[n-1] is the [n] marker in the answer, plus citationUrls (only the sources the text actually cites) and citedDomains.
  • The searches behind the answer - searchQueries lists the web searches Perplexity ran (Pro search), a view no other Perplexity actor gives.
  • Discover in full - all six topics (Top, Tech & Science, Business, Arts & Culture, Sports, Entertainment), about 350-500 stories each, with article text and an average of 22 sources per story.
  • Brand tracking for GEO / AEO - name your brands; every row says how often each one is mentioned and whether its own site is a cited source.
  • Monitor mode - schedule it and get only new or updated stories and answers.
  • No login, no cookies, no browser, no API key. Exports to JSON, CSV, Excel, or straight into your app through the Apify API.

What data can you get from Perplexity?

โ“ Question query๐Ÿ’ฌ Answer answer (markdown) / answerText๐Ÿ”— Sources sources (title, URL, domain, snippet, date)
๐Ÿ”ข sourceCount / citedSourceCount๐ŸŒ citedDomains / citationUrls๐Ÿ’ก relatedQuestions
๐Ÿ”Ž searchQueries (searches Perplexity ran)๐Ÿค– model / modelLabel / mode๐Ÿงต threadTitle / entryIndex / isFollowUp
๐Ÿ‘ค authorUsername / authorName๐Ÿ‘€ viewCount / forkCount๐Ÿ•’ createdAt / updatedAt
๐Ÿ“ฐ Discover title / shortTitle / summary๐Ÿ“Œ bulletSummary (key points)๐Ÿ“„ content / sections (full article)
๐Ÿท๏ธ category / matchedCategories๐Ÿ–ผ๏ธ imageUrl / thumbnailUrl / imageName๐Ÿ“… publishedAt / firstPublishedAt
๐Ÿ“š sourceArticles (card previews)โฑ๏ธ readTimeMinutes / wordCount๐Ÿข brandMentions (per brand)

How to scrape Perplexity with Perplexity Scraper

  1. Create a free Apify account - it comes with monthly free credit.
  2. Open Perplexity Scraper and paste public Perplexity links into Perplexity URLs, or pick Discover topics.
  3. Optionally add keyword filters, a minimum number of sources, a date window or brands to track.
  4. Click Start. The 10-row demo finishes in well under 20 seconds.
  5. Download the results as JSON, CSV or Excel, or read them over the API.

โฌ‡๏ธ Input

ParameterTypeDescription
๐Ÿ” What to scrape
urlsarrayPublic thread URLs (perplexity.ai/search/...), Discover story or Page URLs (perplexity.ai/discover/tech/..., perplexity.ai/page/...), Discover topic URLs, or bare slugs. Optional.
discoverTopicsarraytop, tech, finance, arts, sports, entertainment. Optional; with no URLs and no topics the Top feed is read.
includeFollowUpsbooleanOne row per answer in a thread (default true), or only the first answer.
โœ‚๏ธ Refine results (applied before anything is charged)
sortBystringfeedOrder (default), newest, oldest, mostSources. Any order other than the feed's own reads the whole feed first.
includeKeywords / excludeKeywordsarrayKeep / drop results that mention any of the words. Threads: question + answer. Discover: title, summary, description, key points.
minSourceCountintegerMinimum number of cited sources. Default 0.
publishedAfter / publishedBeforedateDate window on the story's publish date or the answer's creation date.
โž• Extra data
includeArticleTextbooleanDiscover: read each story's article for the full text, sections and every source. Default true, same price.
checkBrandsarrayBrand names to track in every row (brandMentions).
๐Ÿ” Monitoring
monitorModebooleanOnly results not delivered by an earlier run with the same input (or updated since). Default false.
monitorKeystringOptional name for the monitor's memory.
๐Ÿ“ฆ Output
maxResultsintegerMost rows in the run. Default 50, 0 = no limit.
maxResultsPerTopicintegerCap per Discover topic. Default 0 = a fair share of maxResults per topic.
compactbooleanCore fields only, for AI / MCP pipelines. Default false.
โš™๏ธ Advanced
proxyConfigurationobjectNot needed (direct first, residential fallback included in the price). Your own proxy URLs are used first if you add them.
maxConcurrencyintegerParallel requests, default 5.

Input JSON

{
"urls": ["https://www.perplexity.ai/search/europe-has-more-heat-deaths-pe-BDS6xdorS4.4x2WrCC9mAQ"],
"discoverTopics": ["top", "tech"],
"includeKeywords": ["AI", "Nvidia"],
"checkBrands": ["OpenAI", "Nvidia"],
"maxResults": 100
}

A thread link looks like https://www.perplexity.ai/search/<title>-<id> and works when its owner shared it (Share โ†’ public link). A perplexity.ai/search/new?q=... link has no saved answer, and a private, deleted or expired thread cannot be read - those inputs are listed in the run summary and never charged.

Mixing several Discover topics

With several topics and no per-topic cap, each topic gets a fair share of maxResults and the rows come interleaved (1st of each topic, then 2nd...). A story that sits in two picked feeds is delivered once, with both topics in matchedCategories.

Keyword filters match the card on Discover

For Discover stories, keywords are matched against the title, summary, description and key points before the article is read - nothing filtered out is read or charged.

โฌ†๏ธ Output

One row per thread answer (resultType: "threadAnswer") or per Discover story (resultType: "discoverStory"). Real rows from a run on 2026-10-16, shortened:

{
"resultType": "threadAnswer",
"threadUrl": "https://www.perplexity.ai/search/europe-has-more-heat-deaths-pe-BDS6xdorS4.4x2WrCC9mAQ",
"entryIndex": 1,
"query": "Europe has more heat deaths per year than the United States loses to gun deaths.",
"answerText": "Europe Heat Deaths vs. US Gun Deaths: Annual Comparison\nKey Statistics\n\n- Europe: More than 175,000 people die from heat...",
"sources": [{ "position": 1, "title": "Heat claims more than 175,000 lives annually in Europe, latest data ...", "url": "https://news.un.org/en/story/2024/08/1152766", "domain": "news.un.org", "publishedDate": "2024-08-02T00:00:00" }],
"sourceCount": 20,
"citedSourceCount": 8,
"relatedQuestions": ["How do Europe's annual heat deaths compare to US gun-related fatalities"],
"searchQueries": ["Europe heat deaths per year", "United States gun deaths per year", "comparison heat deaths Europe vs gun deaths US"],
"model": "pplx_pro",
"mode": "COPILOT",
"createdAt": "2025-07-16T17:58:15.652003+00:00",
"viewCount": 25816,
"forkCount": 738,
"brandMentions": { "Wikipedia": { "mentionCount": 0, "mentionedInAnswer": false, "citedAsSource": true, "sourcePositions": [16] } }
}
{
"resultType": "discoverStory",
"title": "Anthropic leases first Canadian data center in Nova Scotia",
"summary": "Anthropic will lease the first phase of EverWind's C$10 billion Point Tupper campus for 20 years, with operations slated for late 2027.",
"url": "https://www.perplexity.ai/page/anthropic-leases-first-canadia-1FJVKTWQTCiTyUhPAZrH0A",
"category": "Tech & Science",
"publishedAt": "2026-10-16T22:53:25.712443",
"bulletSummary": ["Anthropic, which makes the Claude chatbot, signed a 20-year lease for a data center in Point Tupper, Nova Scotia, its first in Canada, according to Bloomberg Law."],
"sourceCount": 20,
"sourceDomains": ["theglobeandmail.com", "news.bloomberglaw.com", "everwindfuels.com", "cbc.ca", "globalnews.ca"],
"sources": [{ "position": 1, "title": "Anthropic signs onto $10-billion Nova Scotia data centre project", "url": "https://www.theglobeandmail.com/business/article-anthropic-signs-onto-10-billion-nova-scotia-data-centre-project/", "domain": "theglobeandmail.com", "publishedDate": "2026-10-16T22:13:12" }],
"content": "Anthropic has signed its first data centre lease in Canada, taking capacity at a new campus in Point Tupper, Nova Scotia...",
"wordCount": 417,
"sectionCount": 3,
"readTimeMinutes": 2
}

Fields and measured fill

Measured on 2026-10-16: 239 answers from 27 public threads, and 1,000 Discover stories (6 topics, article text on).

FieldThread answersDiscover stories
query / title, answer / content, model, mode, createdAt / publishedAt100%100%
sources (at least one)39% (answers written without a web search have none)100% (22 per story on average)
source title / url / domain100%100%
source snippet84%27-74%
source publishedDate81%95-96%
relatedQuestions65%-
searchQueries33% (Pro search answers)-
modelLabel98%-
viewCount / forkCount100%-
authorName93%-
bulletSummary, summary, description, imageUrl, sourceDomains-100%
imageWidth / imageHeight-57%
authorImageUrl92%94%

Not shipped, with the reason: likes (Perplexity reports 0 on every answer measured), Discover view/fork counts (0 on every story), and queryCount / totalThreads / articleNumber (the same constant or page index on every card).

Use cases

  • ๐Ÿ“ˆ GEO / AEO teams - see which domains Perplexity cites for the questions in your market and whether your brand is one of them.
  • ๐Ÿงช AI researchers - collect real Perplexity answers with their sources and search steps to study grounding, citation quality and model choice.
  • ๐Ÿ“ฐ Media monitoring - follow Perplexity Discover by topic and get every story with its sources, daily, with monitor mode.
  • ๐Ÿ—‚๏ธ Knowledge teams - archive shared threads your team relies on, follow-ups included, as searchable rows.
  • ๐Ÿ” SEO and content teams - mine relatedQuestions and searchQueries for the questions people and Perplexity actually ask.

How much does it cost to scrape Perplexity?

Pay per result, no start fee: $0.0045 per result ($4.50 per 1,000) - one thread answer or one Discover story with its article text. Private, deleted, filtered, duplicate and unchanged (monitor) results are never charged.

RunResultsCost
Demo (prefill)10$0.045
One Discover topic, default cap50$0.225
Typical100$0.45
Every topic, deep1,000$4.50

Apify's free monthly credit covers about a thousand results.

Integrations, API, scheduling & MCP

  • Schedule a daily run with monitorMode on to receive only new Discover stories or new follow-ups.
  • Webhooks start your own workflow when a run finishes; Make, Zapier, n8n, Google Sheets and Slack integrations work out of the box.
  • API: run Perplexity Scraper from Python or JavaScript with the Apify client.
  • MCP: connect it to Claude, ChatGPT or any MCP client from the MCP tab; compact keeps rows small for AI agents.

More search and news scrapers

Limits & empty runs

  • Existing answers only. Perplexity Scraper reads answers that already exist (public threads, Discover). Perplexity answers every signed-out client's NEW question with "Sign up and repeat your request.", so this actor does not ask new questions.
  • Thread links must be public. In a sample of 340 thread links found on the web, 209 were readable; the rest were private (52), not found (40), expired (16), deleted (7) or malformed (16) - all listed in RUN_SUMMARY, none charged.
  • Discover depth. Each topic feed holds about 350-500 stories at a time (Top: 352, Tech: 498 on 2026-10-16). The feed edition follows the connection's region.
  • Speed. 100 Discover stories with article text in about 15-20 seconds; 1,000 in about 1.5-2 minutes.
  • An empty run (only private or deleted links, or filters that match nothing) is never charged.

โ“ FAQ

How much does Perplexity Scraper cost?

$0.0045 per delivered result, no start fee. A 100-result run is $0.45. Filtered, private, deleted and unchanged results cost nothing.

Can Perplexity Scraper ask Perplexity a new question?

No. Perplexity only answers new questions for signed-in or browser-verified visitors. Perplexity Scraper reads public answers that already exist: shared threads and Discover.

How many results can Perplexity Scraper return per run?

As many as you ask for: every answer of every public thread you paste, and up to about 350-500 stories per Discover topic.

Does Perplexity Scraper return follow-up questions in a thread?

Yes. Each follow-up answer is its own row with entryIndex, entryCount and isFollowUp; switch includeFollowUps off for the first answer only.

Can I track my brand's visibility in Perplexity with Perplexity Scraper?

Yes. Put your brand and competitors in checkBrands; each row reports mentionCount, mentionedInAnswer, firstMentionAtChar, citedAsSource and the sourcePositions of your own site.

Can I schedule Perplexity Scraper to monitor Perplexity Discover?

Yes. Turn on monitorMode and schedule the run; it delivers only stories (or thread answers) that earlier runs did not deliver, or that were updated since.

Can I use Perplexity Scraper with the Apify API or an MCP server?

Yes - call it from any language through the Apify API, or add it to your AI assistant from the MCP tab.

Do I need a Perplexity account, cookies or an API key?

No. Perplexity Scraper reads public pages without a login and needs no Perplexity API key.

Why did my Perplexity Scraper run return fewer results than I asked for?

A thread link may be private, deleted or expired (see RUN_SUMMARY), your filters may have dropped results, or a Discover topic may hold fewer matching stories than maxResults.

Your feedback

Found a bug or missing a field? Open an issue on the Issues tab and we will fix it.