Quora Scraper

Thirdwatch portfolio: 5K users across 88 public Actors, 2M+ records delivered, and >99% run success. Explore all Thirdwatch Actors.

Scrape Quora questions and answers by keyword or by question URL. No login, no cookies, no browser. Give it a few search terms or a list of Quora question links and get back the question (title, follower count, answer count) plus every answer with its full text, author name, author profile URL, author credential, upvotes, comment count, views, shares and dates.

Built for people who need Quora as a data source rather than a website: market researchers mining what customers actually ask, SEO teams looking for question-shaped keywords, founders validating demand, analysts tracking sentiment about a product or brand, and anyone assembling a question-and-answer corpus for LLM training or retrieval.

What you get

Two record types land in the same dataset. Each is billed as one result.

question records

One per question page scraped.

Field	Description
type	Always question
question_id	Quora's numeric question id
question_title	Plain-text question title
question_url	Canonical question URL
question_follower_count	Number of people following the question
question_answer_count	Number of answers written directly to this question
question_total_answer_count	Answers Quora shows on the page including answers merged in from equivalent questions
search_query	The keyword that found this question (null for direct URLs)
search_rank	Position in the keyword results (1 = first)
search_engine	Which discovery engine returned it
scraped_at	ISO 8601 UTC timestamp

answer records

One per answer, up to maxAnswersPerQuestion per question.

Field	Description
type	Always answer
question_id, question_title, question_url	The question this answer was collected from
question_follower_count, question_answer_count	Same counts as the question record, repeated for convenience
answer_id	Quora's numeric answer id
answer_url	Permalink to the answer
answer_text	Full answer as plain text. Paragraphs separated by newlines; list items prefixed with - ; quotes prefixed with >
author_name	Display name, or Anonymous
author_profile_url	https://www.quora.com/profile/... (null for anonymous authors)
author_credential	The credential line Quora shows under the name, e.g. "Software Engineer at Google (2015-present)"
author_is_verified	Quora verified badge
author_is_anonymous	True when the author posted anonymously
is_ai_answer	True for Quora's machine-generated "Assistant" answers (skipped unless includeAiAnswers is on)
is_related_answer	True when Quora surfaced this answer from an equivalent/merged question rather than the page's own question
source_question_title, source_question_url	For related answers: the question it was originally written for
upvotes	Upvote count
comments_count	Number of comments shown on the answer
views	View count
shares	Share count
date	Answer creation time (ISO 8601 UTC)
updated_date	Last edit time (ISO 8601 UTC)
search_query, search_rank, scraped_at	Provenance, as above

Input

{
  "searchQueries": ["best way to learn Python", "is Notion worth it"],
  "questionUrls": ["https://www.quora.com/As-an-intermediate-Python-programmer-how-can-I-take-my-skills-to-the-next-level"],
  "maxQuestionsPerQuery": 10,
  "maxAnswersPerQuestion": 20,
  "includeAiAnswers": false
}

Parameter	Type	Default	Notes
searchQueries	array of strings	[]	Keywords. Each one is resolved to up to maxQuestionsPerQuery Quora question pages
questionUrls	array of strings	[]	Direct question links. Answer permalinks (/answer/Author-Name) are accepted and mapped back to the question. Regional subdomains such as es.quora.com work
maxQuestionsPerQuery	integer	10	1 to 100
maxAnswersPerQuestion	integer	20	0 to 500. Set 0 to collect question records only, which is the cheapest way to turn keywords into a list of Quora questions
includeAiAnswers	boolean	false	Keep Quora's auto-generated "Assistant" answers
proxyConfiguration	object	Apify residential, US	Quora is IP-sensitive. Residential is the safe default; datacenter proxies may work for small runs

You must supply at least one search query or one question URL. Both can be combined in one run; URLs are de-duplicated against search hits.

How it works

Quora renders question pages on the server and embeds the first answers, the question title, follower count and answer count as JSON inside the page. The actor fetches the page over HTTP with a Chrome-shaped TLS fingerprint (no browser is launched), reads that embedded data, and then pages through the remaining answers using the same internal endpoint Quora's own front end calls when you scroll. The endpoint is called anonymously with the per-page form key and the session cookie Quora sets on the first response, so no account, cookie export or login is needed from you.

Quora's on-site search is only available to logged-in users (the search page redirects anonymous visitors to the homepage, and the underlying search API returns nothing for them). Keyword queries are therefore resolved through public web search restricted to quora.com and the matching question pages are then scraped directly. In practice this returns the same high-traffic questions that Quora's own search would rank first, but the ordering is the search engine's, not Quora's, and the Question / Answer / Profile / Space type filters of Quora's native search are not available.

Each question costs one page request plus roughly one request per 8-10 additional answers. A 2-3 second pause is kept between requests and 429/503 responses back off exponentially. A blocked question page is retried once on a fresh proxy session before being counted as blocked.

Why answers may be fewer than the question's answer count

• question_answer_count counts answers written directly to that question, while Quora's page also mixes in answers from equivalent questions; the mixed list is what gets scraped and each record says which case it is via is_related_answer.
• Quora's AI "Assistant" answers are skipped unless you enable includeAiAnswers.
• Collapsed, deleted, or moderation-hidden answers are not served to anonymous visitors.
• maxAnswersPerQuestion caps the count.

Zero results and failures

The run stays honest about what happened:

• Keyword found no Quora questions: the run succeeds with a warning and no records for that keyword.
• Question URL returns 404 or bounces to the homepage (removed or merged question): counted as not found, run continues.
• Quora or the search engines could not be reached for any target: the run fails with a status message instead of quietly succeeding with an empty dataset, so a scheduled run does not silently degrade.

Use cases

Keyword and content research. Quora questions are literal user intent. Run your seed topics through searchQueries with maxAnswersPerQuestion: 0 to harvest hundreds of question-shaped keywords with follower counts as a demand signal, then scrape the answers for the ones worth writing about.

Product and market research. Search a category ("best CRM for small business", "alternatives to X") and read what real users recommend, complain about and compare. upvotes and views tell you which opinions the community endorses.

Brand and reputation monitoring. Schedule daily runs on your brand and competitor names. New answers appear with their date, author and credential, so you can respond or escalate.

Expert sourcing. Filter answers by author_credential and upvotes to find practitioners in a niche, with profile URLs ready for outreach.

LLM and RAG datasets. Clean question/answer pairs with engagement signals and timestamps, ready to embed. Answers are plain text, not HTML.

Tips

• Start with maxAnswersPerQuestion: 0 to preview which questions a keyword returns before paying for answers.
• Very long-tail keywords sometimes return nothing; broaden the phrasing or drop qualifiers.
• If you already know the question URLs (from Google, a sitemap, a previous run), pass them directly; this skips discovery and is faster.
• For thousands of questions, split across multiple runs or raise the run timeout; each question takes about 3-8 seconds including pagination.

Integrations

Runs on the Apify platform, so you can trigger it via API, schedule it, chain it with other actors, and export the dataset as JSON, CSV, Excel or push it to Google Sheets, Airtable, Zapier, Make and webhooks.

curl -X POST "https://api.apify.com/v2/acts/thirdwatch~quora-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"searchQueries": ["best way to learn Python"], "maxQuestionsPerQuery": 5, "maxAnswersPerQuestion": 10}'

Legal

This actor collects only publicly visible content that Quora serves to anonymous visitors. Answers and questions are user-generated and belong to their authors; make sure your use complies with Quora's terms, applicable copyright law and data-protection rules (GDPR, CCPA) for any personal data such as author names.

Other Thirdwatch actors you may need

Reddit Scraper for the other big discussion source, Google Search Scraper for SERP data, Trustpilot Scraper and G2 Scraper for structured reviews. See all Thirdwatch actors.

Output fields

• One row per result with stable keys.

Limitations

• Results reflect the public source at run time; removed or restricted items disappear between runs.
• Fields that are not publicly disclosed stay empty rather than being guessed.
• Use public data lawfully and respect applicable privacy and platform rules.

Last verified: 2026-09

More scrapers at thirdwatch.dev.