Stack Overflow Scraper — Questions, Scores & Tags

Search Stack Overflow — or any Stack Exchange site — and get one row per question: score, answer count, views, tags, author and reputation, accepted-answer flag, and ISO dates.

No API key required. $0.005 per question — a search that matches nothing is recorded and not charged.

Input

{ "query": "async await", "maxQuestions": 25, "sort": "votes" }

By tag, on another site, with a score floor:

{
  "tagged": "python;pandas",
  "site": "stackoverflow",
  "minScore": 50,
  "answeredOnly": true,
  "maxQuestions": 100
}

Field	Type	Notes
query	string	Free-text search over titles and bodies.
queries	string[]	Up to 20 searches per run.
tagged	string	Semicolon-separated tags, e.g. python;pandas. Works with or without a query.
site	string	stackoverflow, serverfault, superuser, askubuntu, math, unix, dba…
sort	enum	votes, activity, creation, relevance.
maxQuestions	integer	1–100 per search. You are charged per question returned.
answeredOnly	boolean	Only questions with an accepted answer.
minScore	integer	Skip questions below this score.
apiKey	string	Optional — see Limits.

Output

{
  "ok": true,
  "questionId": 37576685,
  "title": "Using async/await with a forEach loop",
  "link": "https://stackoverflow.com/questions/37576685/...",
  "score": 3353,
  "answerCount": 35,
  "viewCount": 2481265,
  "isAnswered": true,
  "hasAcceptedAnswer": true,
  "tags": ["javascript", "node.js", "promise", "async-await"],
  "author": "Saad",
  "authorReputation": 52193,
  "createdAt": "2016-06-01T18:55:58.000Z",
  "lastActivityAt": "2026-08-14T09:12:40.000Z",
  "licence": "CC BY-SA 4.0"
}

What it gets right

• Dates are ISO-8601. Stack Exchange returns epoch seconds; a bare 1464807358 is not a date and will not sort in a spreadsheet.
• One row per question, so filtering, sorting and importing all just work.
• site reaches the whole network — Server Fault, Super User, Ask Ubuntu, Math, Unix and the rest, not just Stack Overflow.
• API errors are surfaced, not swallowed. Stack Exchange reports a bad site or a bad parameter in a 200-shaped body with an error_message; that message reaches you instead of an empty run.
• Numbers are numbers and flags are booleans — score, answerCount, viewCount, isAnswered, hasAcceptedAnswer.

Pricing

$0.005 per question returned. No start fee. A search matching nothing produces a questionCount: 0 record and is not charged, so probing a narrow tag is free.

Limits

⚠️ The keyless API allows 300 requests per day per IP, and Apify runs from shared addresses, so that pool is not yours alone. This Actor reads the remaining quota from every response, reports it as quotaRemaining, and when it runs out it writes an honest uncharged record rather than a run of empty results.

Pass your own free key via apiKey to raise the limit to 10,000/day — register at stackapps.com. It takes a minute and costs nothing.

• 20 searches per run, up to 100 questions each.
• Returns question metadata, not answer bodies.
• Content is CC BY-SA; the licence field carries the exact version per question.

FAQ

Do I need an API key? No — but supply one if you are running large or frequent jobs.

Can I get the answers themselves? Not in this Actor; it returns question metadata including answer counts and the accepted-answer flag.

Which sites work? Any Stack Exchange site — pass its API slug in site.

Changelog

• 0.1 (2026-09-10) — first release. One row per question, whole Stack Exchange network, tag and score filters, bulk searches, quota surfaced with optional caller key.