Stack Overflow Data & API Alternative | $0.60 / 1K

Stack Overflow data export: questions and answers by tag or keyword, for any Stack Exchange site

Get Stack Overflow data (and data from any other Stack Exchange site) as flat rows you can open in a spreadsheet, load into a database or feed to an LLM. The Actor reads the official Stack Exchange API: no login, no browser, no scraping of pages. Filter by tag, keyword or both, optionally pull every answer, and switch on incremental mode to get only the questions posted since your last run.

$0.60 per 1,000 rows. No start fee. 200 questions cost about 12 cents.

What people use it for

• Stack Overflow questions by tag: export the top-voted, newest or most active questions for a tag such as python, react or kubernetes.
• Stack Overflow Q&A dataset: questions plus every answer, one row each, with scores and accepted-answer flags, for LLM fine-tuning data, RAG knowledge bases and research.
• Monitor new questions about your product or library: run incremental mode on a schedule and send new questions to Slack, a sheet or your support queue. Developer-relations and support teams use this to find people who are stuck.
• Stack Exchange data export: change the site input to serverfault, superuser, askubuntu, stats (Cross Validated) or any other Stack Exchange API site name.
• Content and SEO research: find which questions in your topic get views and votes, and what the accepted answers look like.
• Developer trend analysis: track tag activity and question volume over time.

Why use this instead of calling the Stack Exchange API yourself?

The official API is free, documented and a good choice if you want to write the code. This Actor saves the plumbing:

• Flat rows with full bodies (bodyHtml), tags, scores, view counts, owner reputation and direct links, ready for CSV, Excel or JSON export.
• Answers in batches: up to 100 questions per request, instead of one request per question.
• Incremental mode: "only what's new since last time", tracked for you in a key-value store in your own account. It uses the API's own fromdate filter, so the server does the filtering.
• Scheduling, webhooks and integrations from Apify, with no server to maintain.

Things it does not do: no comments, no user profile export, no tag statistics, no unanswered-only filter, and no body text in Markdown (bodies are HTML as the API returns them).

Incremental mode

Set mode to changes. The first run takes questions created since firstRunSince (30 days by default, all for a deliberate backfill), and every later run returns only newer ones for that exact tag, query and site combination. A run cut short by maxItems still advances its mark from what it actually emitted, so the next run does not repeat rows or skip any. Answers are not tracked separately: a later run returns answers only for the new questions it reads.

The API quota and the optional key

Stack Exchange gives anonymous callers a shared 300 requests per day. One request returns up to 100 rows, so that covers a lot of rows. If you run often or track many tags, register a free key at stackapps.com (instant) and paste it into Stack Apps API key. The same account's quota rises to 10,000 requests per day.

Input

Input	Default	Notes
site	stackoverflow	Any Stack Exchange site API name, e.g. serverfault, superuser, askubuntu
tags	none	AND-matched. Leave empty to use query alone
query	none	Full-text search, combinable with tags
mode	all	all or changes (incremental)
sort	creation	creation, activity, votes, hot. Incremental mode always uses creation
includeAnswers	false	See the note below
minScore	0	Drops questions and answers scoring lower. Applied after the rows are fetched, so on a huge tag a high minScore can read many pages of quota; narrow the tags or add a key
maxItems	200	Cost cap, counts questions and answers together
firstRunSince	30 days	changes mode only
apiKey	none	Optional, see above

Answers and maxItems: the Actor saves all matching questions first, then fetches answers, and maxItems counts both. To get answers, choose a filter narrow enough that questions plus answers fit under maxItems. If the questions alone reach the cap, you get no answers.

Example input, top-voted questions on another Stack Exchange site:

{ "site": "askubuntu", "sort": "votes", "maxItems": 30 }

Example input, a small Q&A dataset on a niche tag:

{ "tags": ["polars"], "sort": "votes", "includeAnswers": true, "maxItems": 150 }

Example input, new questions for a tag since the last run (schedule it):

{ "tags": ["rust"], "mode": "changes", "firstRunSince": "7 days", "maxItems": 100 }

Output

One row per question, plus one row per answer if you turn answers on. Real example (question row):

{
  "questionId": 17823,
  "title": "How to list all installed packages",
  "bodyHtml": "<p>I'd like to output a list of all installed packages into a text file ...</p>",
  "tags": ["apt", "package-management"],
  "score": 2749,
  "viewCount": 6066264,
  "answerCount": 26,
  "isAnswered": true,
  "acceptedAnswerId": 17829,
  "ownerName": "Ivan",
  "ownerUserId": 2390,
  "ownerReputation": 59195,
  "createdUtc": 1292551729,
  "createdAtIso": "2010-12-17T02:08:49.000Z",
  "link": "https://askubuntu.com/questions/17823/how-to-list-all-installed-packages"
}

Answer rows carry answerId, questionId, bodyHtml, score, isAccepted, owner fields, createdAtIso and link.

Pricing: pay per row, worked examples

Event	Price	When
question-scraped	$0.0006	One question row
answer-scraped	$0.0006	One answer row

Run	Rows	Cost
30 top questions for a tag	30	about $0.02
Default run (maxItems 200)	200	$0.12
60 questions with their answers (about 115 rows)	115	about $0.07
10,000 rows	10,000	$6

You pay only for rows saved, so maxItems is a hard cost cap. Platform usage is included in the per-row price.

Integrations and scheduling

• Apify Schedules: run incremental mode hourly or daily for ongoing monitoring.
• Webhooks, Make, n8n and Zapier: send new questions to Slack, email, Notion, Jira or your CRM.
• Google Sheets: export datasets to a sheet.
• API and clients (JavaScript, Python): start runs and read the dataset from code. AI agents can use it through Apify's MCP server.

FAQ

Is there a Stack Overflow API alternative that needs no key? This Actor uses the official Stack Exchange API and works without a key, within the shared 300 requests per day. A free key lifts that to 10,000.

Can I export Stack Overflow questions to CSV or Excel? Yes. Run the Actor and download the dataset as CSV, Excel, JSON or HTML table.

Can I get answers too? Yes, set includeAnswers. Keep the filter narrow, because maxItems is shared between questions and answers.

Can I get data from other Stack Exchange sites? Yes, set site to the site's Stack Exchange API name, for example serverfault, superuser, askubuntu or stats (Cross Validated).

How do I get only new questions? Use mode: changes and schedule it.

Does it download the whole Stack Overflow archive? No, that is what Stack Exchange's public data dump is for. This Actor is for targeted tag or keyword pulls and monitoring.

Is it legal? It reads the official public API within its terms and quotas. Content on Stack Exchange is licensed CC BY-SA, so keep attribution (the link field) if you republish it.

What are alternatives? sian.agency/stack-exchange-scraper covers Stack Overflow and the Stack Exchange network, including URL and id lookups, and sheshinmcfly/stackoverflow-scraper is a Q&A Actor. Compare current pricing and features on their pages.

Related Actors

• Hacker News Data & API Alternative: stories, comments and a firehose for new items.
• Reddit Scraper & API Alternative: subreddits, search and comments with incremental mode.
• Y Combinator Companies Scraper: the YC directory as data.
• Google Trends Scraper: compare search interest for technologies.
• sian.agency/stack-exchange-scraper: third-party Actor for Stack Exchange with URL lookups.