Stack Overflow Q&A Scraper — Questions, Answers, Comments & Markdown

Extract questions with their best answers from Stack Overflow and 30+ Stack Exchange sites through the official API. Search by tags or keywords, get the top-voted or hottest questions, or pass specific question URLs. Every question comes with its answers (the accepted one first), in HTML and Markdown, plus votes, comments, closed status and quality scores — clean input for AI training data, RAG pipelines and knowledge bases. No login, no account.

🧵 Complete threads	Each question with its accepted answer and top answers — none left out
📝 Markdown + plain text	Bodies as HTML, Markdown and plain text, ready for LLMs
💬 Comments	Comments on questions and answers (optional)
🔎 Filters that work	Minimum score, maximum answers, dates, answered only
🔗 Question URLs	Pull specific questions from any Stack Exchange site

Data you get

Per question

Field	Description
questionId / site / url	Identity
title / body / bodyMarkdown / bodyText	Question as HTML, Markdown and plain text
tags	Tags
score / upVotes / downVotes / favoriteCount / viewCount	Engagement
answerCount / commentCount / isAnswered / acceptedAnswerId	Discussion
isClosed / closedReason / closedDate	Closed status
bountyAmount / bountyClosesDate	Open bounty
author / authorReputation / authorId	Asker
createdAt / lastEditDate / lastActivityDate / protectedDate / contentLicense	Dates and license
answers	Accepted answer first, then the highest voted (see below)
comments	Comments on the question (optional)
linkedQuestions	Questions linked to this one (optional)
qualityScore / frustrationScore / discussionDepth	0–100 quality, frustration signals and discussion depth

Per answer — answerId, author, authorReputation, authorId, score, upVotes, downVotes, isAccepted, body (HTML), bodyMarkdown, createdAt, lastEditDate, commentCount, comments (optional), url

Per comment — commentId, author, score, body, createdAt

Popular tags mode — name, count, hasSynonyms, isRequired

Quick start

Top Python questions with their best answers

{ "mode": "tags", "tags": ["python"], "maxResults": 200, "maxAnswersPerQuestion": 3 }

Keyword search

{ "mode": "search", "keywords": "docker compose volumes", "sortBy": "relevance", "maxResults": 50 }

Well-vetted recent questions

{ "mode": "tags", "tags": ["rust"], "sortBy": "creation", "fromDate": "2026-01-01", "minScore": 5, "maxResults": 100 }

Specific questions with comments

{ "questionUrls": ["https://stackoverflow.com/questions/231767", "https://math.stackexchange.com/questions/1"], "includeComments": true }

How to use

1. Pick a mode: by tags, keyword search, top questions, popular tags — or paste question URLs.
2. Choose the site (Stack Overflow, Super User, Mathematics, Cross Validated…) and the sort order.
3. Narrow it down with dates, minimum score, maximum answers and only answered.
4. Choose how many answers per question, and turn on comments or linked questions if you need them.
5. Run. The status message shows how many questions were returned out of how many match.

Use cases

• AI training data — problem/solution pairs with the accepted answer, in Markdown.
• RAG pipelines — a knowledge base for a technology stack, with votes to rank sources.
• Developer research — what people struggle with, how fast questions get answered, which get closed.
• Content and documentation gaps — high-view questions with low-scored answers.
• Community analytics — tags, activity and engagement across Stack Exchange sites.

Input

Field	Type	Description
mode	select	tags, search, top, popularTags or questions
tags	array	Tags to filter by (tags mode)
keywords	string	Search text (search mode)
questionUrls	array	Question URLs or IDs, any Stack Exchange site
site	select	Stack Exchange site
sortBy	select	Votes, activity, newest, hot, week, month (search also: relevance)
onlyAnswered	boolean	Only questions with an accepted or upvoted answer (default true)
minScore	integer	Minimum question score
maxAnswers	integer	Maximum number of answers a question can have (0 = no limit)
fromDate / toDate	date	Creation date range
includeAnswers	boolean	Answers per question (default true)
maxAnswersPerQuestion	integer	Accepted first, then highest voted (default 3)
includeComments	boolean	Comments on questions and answers
includeLinked	boolean	Linked questions
maxResults	integer	Max questions (or tags) — up to 2,500 per run
apiKey	string	Optional Stack Apps key: lifts the 2,500 limit per query

Output example

{
  "questionId": "16047306",
  "site": "stackoverflow",
  "title": "How is Docker different from a virtual machine?",
  "bodyMarkdown": "I keep rereading [the Docker documentation](https://docs.docker.com/) to try to understand the difference between Docker and a full VM…",
  "tags": ["docker", "virtual-machine"],
  "score": 4501,
  "upVotes": 4524,
  "downVotes": 23,
  "viewCount": 965917,
  "answerCount": 27,
  "isAnswered": true,
  "isClosed": false,
  "author": "zslayton",
  "createdAt": "2013-04-16T21:19:47.000Z",
  "url": "https://stackoverflow.com/questions/16047306/how-is-docker-different-from-a-virtual-machine",
  "answers": [
    {
      "answerId": 16048358,
      "author": "Ken Cochrane",
      "score": 4045,
      "isAccepted": true,
      "bodyMarkdown": "Docker originally used [LinuX Containers](https://linuxcontainers.org/) (LXC)…",
      "url": "https://stackoverflow.com/a/16048358"
    }
  ],
  "qualityScore": 92
}

Performance

Measured runs:

Request	Questions	Time
Keyword search, 30 questions + answers	30	~5 s
100 questions + 3 answers each	100	~10–15 s
10 questions + 5 answers + all comments	10	~7 s
20 questions + linked questions	20	~13 s
500 questions + answers	500	~25 s
1,500 questions without answers	1,500	~10 s

Pricing

Pay per result: $0.002 per question (or tag in popular tags mode), plus $0.00005 per run start. Answers, comments and linked questions are included in the question price.

Questions	Price
1,000	$2.00
10,000	$20.00

FAQ

Do I need an API key? No. Runs work without one. A free key from stackapps.com only lifts the 2,500-results-per-query limit.

Which answers are included? The accepted answer first, then the highest-voted ones, up to your maximum per question.

Can I get unanswered questions? Yes — turn off only answered; combine with maximum answers to target questions with few answers.

Does it work on other Stack Exchange sites? Yes — choose the site, or paste question URLs from any Stack Exchange site.

Keywords

stackoverflow scraper, stack overflow api, stack exchange scraper, stack overflow answers, programming q&a dataset, ai training data, rag dataset, llm fine-tuning data, developer questions, stack overflow markdown

More actors

Browse all my scrapers on my Apify profile.

Disclaimer

This actor collects only publicly available content through the official Stack Exchange API and is not affiliated with Stack Exchange. Stack Exchange content is licensed under CC BY-SA; you are responsible for attribution and for how you use the data.

Changelog

Version	Date	Changes
1.1	2026-10-29	New: answers in Markdown with votes, the accepted answer always first; comments on questions and answers; question URLs from any Stack Exchange site; plain-text bodies, up/down votes, closed status and bounties; up to 2,500 questions per run. Fixed: results stopped after the first page (~100); minimum score and maximum answers filters now work for every sort order; every answered question now comes with its answers; runs no longer end empty when the daily API limit is reached; HTML entities in titles and names decoded. Faster runs and a summary in the status message.
1.0	2026-04-21	Initial release