💬 Quora Scraper — Questions & Answers

$5/1K Answers 💰 — View on Apify Store

Scrape questions and answers from Quora: question title, topic, answer count, followers — plus full answers with author, author bio, upvotes and publish date.

Works from direct question URLs only. Quora's /search route redirects anonymous sessions to the sign-in page — measured over a residential IP and through a real browser, both land on "Continue with Google". Search keywords therefore cannot work without an account, and this Actor will not pretend otherwise. Pass question URLs in urls.

Quora is one of the most anti-bot platforms (Cloudflare + login walls + lazy loading). This actor was built to survive it: Apify RESIDENTIAL proxy by default, cookie seeding, stealth init script and lazy-load scrolling.

---

🚀 How it works

1. Give it question URLs (urls).
2. searchQueries is still accepted for backward compatibility, but Quora gates /search behind a login — those runs end with a diagnostic record and are not charged.
3. For each question it loads the page (cookie-seeded residential session), scrolls to trigger lazy loading, expands answers (optional), and extracts:

{
  "url": "https://www.quora.com/What-is-the-best-way-to-learn-programming",
  "question": "What is the best way to learn programming?",
  "topic": "Programming",
  "upvotes": 1200,
  "answers_count": 350,
  "followers": 9800,
  "answers": [
    {
      "author": "Jane Doe",
      "author_bio": "Software Engineer at X",
      "content": "The best way is to build things...",
      "upvotes": 452,
      "is_expanded": true,
      "published_at": "3y"
    }
  ],
  "scrapedAt": "2026-08-13T02:55:00.000Z"
}

Missing fields are returned as null — honest data, never invented. Fields that Quora does not show (e.g. followers on some pages) stay null.

---

⚙️ Input

Field	Type	Default	Description
urls	stringList	—	Quora question URLs to scrape
searchQueries	stringList	—	Not working: Quora redirects /search to the sign-in page for anonymous sessions. Accepted so old task configurations do not break; produces a diagnostic, never a charge.
maxQuestionsPerSearch	integer	5	Max question links extracted per search query (1–20)
maxAnswersPerQuestion	integer	5	Max answers extracted per question (1–20). Values above 4 use Quora's own paginated query — see note below
expandAnswers	boolean	false	Click "Continue reading" so long answers are extracted in full
proxyConfig	proxy	RESIDENTIAL	Quora blocks datacenter IPs — keep RESIDENTIAL

urls is required in practice — searchQueries alone cannot return data (see above).

About maxAnswersPerQuestion above 4 — the Quora question page itself only carries 4 answers (measured across 6 fetches of 3 different questions). Everything past that is fetched with Quora's own paginated GraphQL query, which requires a real browser session (it sends a Quora-Formkey header plus session cookies). Measured with that route: 20 answers per question on 4 different questions. When the remote-browser endpoint is not configured, the run delivers the 4 answers the page carries and says so in the log — it never pads the output to look like it delivered more.

---

📦 Output

Each dataset item is one question with its answers nested. Field reference:

Field	Type	Notes
url	string	Question page URL (always present)
question	string	Question title (always present)
topic	string | null	Primary topic
upvotes	integer | null	Question upvotes. Always null: the page payload carries no upvote count for the question itself (checked across 48 data blobs). Answer upvotes below are real.
answers_count	integer | null	Answers count shown on page
followers	integer | null	Question followers
answers[]	array	See below
answers[].author	string | null	Author name
answers[].author_bio	string | null	Author credential/bio line
answers[].content	string | null	Answer text
answers[].upvotes	integer | null	Answer upvotes
answers[].is_expanded	boolean	true = full text loaded
answers[].published_at	string | null	Date as displayed on Quora (e.g. "3y", "2 weeks ago")
scrapedAt	string	ISO timestamp

---

💰 Pricing & free tier

• $5 per 1K answers extracted (PPE: answer-scraped). Questions themselves cost nothing.
• Quora serves, on the same question page, answers written under equivalent or merged questions — and presents them as answers to that question. They are delivered and billed like any other answer, and each one carries is_equivalent_answer: true plus source_question_url so you can tell exactly which question it came from.
• Free tier: limited to 1 question and 2 answers per run (so you can try the actor for free). Upgrade your Apify plan to unlock full limits.
• Diagnostic runs (login wall / Cloudflare challenge / zero results) are not charged.

---

🛡️ Anti-bot notes (honest limitations)

• Quora aggressively blocks datacenter IPs. Always run with Apify RESIDENTIAL proxy (the default).
• If a run returns a diagnostic item with setup_status: "DIAGNOSTIC_GUIDE", the IP was blocked or the page hit a login wall. Wait ~30 minutes and retry.
• If expandAnswers: false, long answers may be truncated by Quora (extracted text stops at "Continue reading"). Enable expandAnswers for full content.
• published_at is captured as Quora displays it (relative like "3y" or "2 weeks ago") — Quora does not expose exact timestamps in the public DOM.
• Answers are read from the page's own data payload, not from CSS classes, so obfuscated class names do not affect it.
• A question page carries only its first batch of answers — measured 4 per page across 6 fetches of 3 different questions. Asking for more returns what the page has; the rest needs Quora's paginated API, which is not implemented.
• Quora mixes in answers written under equivalent/merged questions. Measured: 8 of 12 answers came from a related question rather than the one requested. Every answer carries is_equivalent_answer and source_question_url so you can tell them apart — and you are only charged for answers belonging to the question you asked for.

---

🔍 Use cases

• Content research / SEO: find the most-upvoted answers on a topic
• Competitor analysis: what questions your audience asks
• Building a Q&A knowledge base for RAG / training data
• Influencer discovery: top answer authors on a topic (via author + author_bio)

---

Scrape responsibly. Respect Quora's Terms of Service and rate limits. This actor is not affiliated with Quora.

---

💬 Quora Scraper — Perguntas & Respostas

$5/1K Respostas 💰 — Veja na Apify Store

Extraia perguntas e respostas do Quora: título da pergunta, tópico, upvotes, número de respostas, seguidores — e as respostas completas com autor, bio do autor, upvotes e data de publicação. Funciona com URLs diretas de perguntas ou palavras-chave de busca.

O Quora é uma das plataformas mais anti-bot (Cloudflare + login walls + lazy loading). Este actor foi construído para sobreviver: proxy RESIDENTIAL da Apify por padrão, cookie seeding, script de stealth e scroll para lazy loading.

---

🚀 Como funciona

1. Informe URLs de perguntas (urls). O campo searchQueries continua aceito por compatibilidade, mas o Quora exige login na busca: esses runs terminam com diagnóstico e não são cobrados.
2. Para palavras-chave, o actor visita https://www.quora.com/search?q=... e descobre os links de perguntas automaticamente.
3. Para cada pergunta, carrega a página (sessão residencial com cookie seed), rola para ativar o lazy loading, expande respostas (opcional) e extrai:

{
  "url": "https://www.quora.com/What-is-the-best-way-to-learn-programming",
  "question": "What is the best way to learn programming?",
  "topic": "Programming",
  "upvotes": 1200,
  "answers_count": 350,
  "followers": 9800,
  "answers": [
    {
      "author": "Jane Doe",
      "author_bio": "Software Engineer at X",
      "content": "The best way is to build things...",
      "upvotes": 452,
      "is_expanded": true,
      "published_at": "3y"
    }
  ],
  "scrapedAt": "2026-08-13T02:55:00.000Z"
}

Campos ausentes retornam null — dados honestos, nunca inventados. Campos que o Quora não exibe (ex.: seguidores em algumas páginas) permanecem null.

---

⚙️ Input

Campo	Tipo	Padrão	Descrição
urls	stringList	—	URLs de perguntas do Quora para extrair
searchQueries	stringList	—	Não funciona: o Quora redireciona /search para a página de login em sessão anônima. Continua aceito para não quebrar configurações antigas; gera diagnóstico, nunca cobrança.
maxQuestionsPerSearch	integer	5	Máx. de links de perguntas por busca (1–20)
maxAnswersPerQuestion	integer	5	Máx. de respostas por pergunta (1–20). Acima de 4 usa a query paginada do próprio Quora — veja a nota abaixo
expandAnswers	boolean	false	Clica em "Continue reading" para respostas longas completas
proxyConfig	proxy	RESIDENTIAL	O Quora bloqueia IPs de datacenter — mantenha RESIDENTIAL

Na prática, urls é obrigatório — searchQueries sozinho não devolve dado (veja acima).

Sobre maxAnswersPerQuestion acima de 4 — a página de pergunta do Quora carrega apenas 4 respostas (medido em 6 leituras de 3 perguntas distintas). O que passa disso vem da query paginada do próprio Quora, que exige sessão real de navegador (manda o cabeçalho Quora-Formkey mais os cookies da sessão). Medido por essa rota: 20 respostas por pergunta em 4 perguntas distintas. Sem o endpoint de navegador remoto configurado, o run entrega as 4 respostas que a página traz e registra isso no log — nunca infla a saída para parecer que entregou mais.

---

📦 Output

Cada item do dataset é uma pergunta com suas respostas aninhadas. Referência:

Campo	Tipo	Notas
url	string	URL da pergunta (sempre presente)
question	string	Título da pergunta (sempre presente)
topic	string | null	Tópico principal
upvotes	integer | null	Upvotes da pergunta
answers_count	integer | null	Nº de respostas exibido na página
followers	integer | null	Seguidores da pergunta
answers[]	array	Ver abaixo
answers[].author	string | null	Nome do autor
answers[].author_bio	string | null	Bio/credencial do autor
answers[].content	string | null	Texto da resposta
answers[].upvotes	integer | null	Upvotes da resposta
answers[].is_expanded	boolean	true = texto completo carregado
answers[].published_at	string | null	Data como exibida no Quora (ex.: "3y", "2 weeks ago")
scrapedAt	string	Timestamp ISO

---

💰 Preço & free tier

• $5 por 1K respostas extraídas (PPE: answer-scraped). Perguntas não custam nada.
• O Quora serve, na mesma página da pergunta, respostas escritas sob perguntas equivalentes ou mescladas — e as apresenta como respostas daquela pergunta. Elas são entregues e cobradas como qualquer outra, e cada uma traz is_equivalent_answer: true e source_question_url para você saber exatamente de onde veio.
• Free tier: limitado a 1 pergunta e 2 respostas por execução (teste grátis). Faça upgrade do plano Apify para liberar os limites completos.
• Execuções de diagnóstico (login wall / Cloudflare / zero resultados) não são cobradas.

---

🛡️ Notas anti-bot (limitações honestas)

• O Quora bloqueia agressivamente IPs de datacenter. Sempre execute com proxy RESIDENTIAL da Apify (padrão).
• Se uma execução retornar um item de diagnóstico com setup_status: "DIAGNOSTIC_GUIDE", o IP foi bloqueado ou a página caiu em login wall. Aguarde ~30 minutos e tente novamente.
• Com expandAnswers: false, respostas longas podem vir truncadas pelo Quora (o texto para em "Continue reading"). Ative expandAnswers para conteúdo completo.
• published_at é capturado como o Quora exibe (relativo, ex.: "3y" ou "2 weeks ago") — o Quora não expõe timestamps exatos no DOM público.
• Os nomes de classe do Quora são ofuscados; a extração é estrutural (link de perfil + botão de upvote = card de resposta), o que é robusto, mas pode ocasionalmente deixar passar páginas com formatação incomum.

---

🔍 Casos de uso

• Pesquisa de conteúdo / SEO: respostas mais votadas sobre um tema
• Análise de concorrentes: quais perguntas seu público faz
• Base de conhecimento Q&A para RAG / dados de treinamento
• Descoberta de influenciadores: principais autores de um tópico (via author + author_bio)

---

Colete com responsabilidade. Respeite os Termos de Serviço e limites de taxa do Quora. Este actor não é afiliado ao Quora.