Stack Overflow Scraper: Monitor Tags, Keywords, and New Questions

Scrape Stack Overflow questions by tag, keyword, or search query. Export titles, bodies, tags, authors, scores, and timestamps to JSON, CSV, or Excel. Deduped across runs so you only see new questions. No login. No API key. Pay per question.

Keywords this actor is built for: Stack Overflow scraper, Stack Overflow API data, scrape Stack Overflow questions, Stack Overflow tag monitor, Stack Exchange scraper, developer lead generation, Stack Overflow JSON export.

---

What you get in 30 seconds

flowchart LR
    A[Your tags<br/>python, langchain] --> B[Actor polls<br/>Stack Exchange API]
    B --> C[Filter by keyword,<br/>score, age, answered]
    C --> D[Deduped JSON feed<br/>of new questions only]

Paste a tag. Pick a filter. Get a clean JSON feed of new Stack Overflow questions every run. That is the whole product.

---

Who this Stack Overflow scraper is for

You are a...	You use this to...
Devtool founder	Find every "how do I use X" question in your category and turn it into warm outbound
DevRel engineer	Build a queue of unanswered questions in your tag, ready for a helpful reply
Technical marketer	Mine real searcher language for blog posts, docs, and landing pages
Support team	Catch user pain on Stack Overflow before it hits GitHub issues or Twitter
Product manager	Track tag volume week over week to see when a category is heating up

---

How it works

flowchart TD
    A[Input: tags,<br/>search queries, keywords] --> B[Stack Exchange API v2.3]
    B --> C[Paginate 100 per page]
    C --> D{Filters}
    D --> E[Keyword match]
    D --> F[Score floor]
    D --> G[Age window]
    D --> H[Unanswered only]
    E & F & G & H --> I[Push to Apify dataset]
    I --> J[Store ID in<br/>SEEN_IDS key value store]
    J -.->|Next run| K[Skip seen IDs]

1. You pass tags, search queries, or both.
2. Actor calls the official Stack Exchange API v2.3 with pagination.
3. Results are filtered by your keyword list, score floor, age window, and answered state.
4. Matching questions get pushed to the dataset.
5. Every question ID is stored in a named key value store so future runs skip duplicates.

Schedule it every hour on Apify Scheduler and you get a deduped stream of new questions. Nothing else.

---

Quick start

Watch the langchain tag for new questions about vectors or embeddings:

{
  "tags": ["python", "langchain"],
  "keywords": ["vector", "embedding"],
  "maxAgeHours": 168,
  "onlyUnanswered": true
}

Outbound list for postgres performance pain:

{
  "tags": ["postgresql", "performance"],
  "searchQueries": ["postgres slow query"],
  "onlyNoAcceptedAnswer": true,
  "maxAgeHours": 72,
  "maxQuestionsTotal": 100
}

DevRel answer queue on a sibling Stack Exchange site:

{
  "tags": ["kubernetes"],
  "site": "devops.stackexchange",
  "onlyUnanswered": true
}

Or run it from the command line:

curl -X POST "https://api.apify.com/v2/acts/scrapemint~stackoverflow-lead-monitor/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"tags":["langchain"],"keywords":["vector"],"onlyUnanswered":true}'

---

This scraper vs the alternatives

flowchart LR
    M[Manual tag checking] --> X[20 questions per day,<br/>miss 80% of volume]
    S[SaaS intent tools] --> Y[$99 to $499 per month,<br/>dashboard only]
    A[This Apify actor] --> Z[Pay per question,<br/>raw JSON, your webhook]

Feature	Manual checking	SaaS tools	This actor
Pricing	Free, but costs time	$99 to $499 per month	Pay per question, first 50 per run free
Tag limit	Unlimited, if you click them all	10 to 50 per tier	Unlimited
Unanswered filter	Sort by no answers	Premium tier only	Built in
Other Stack Exchange sites	Tab hopping	Stack Overflow only	Any site (DBA, ServerFault, Super User)
Scheduling	You	Hourly	Every 10 minutes
Dedup across runs	Your memory	Yes, vendor owned	Yes, in your key value store
Output	Browser tab	Dashboard or CSV	JSON, CSV, Excel, webhook, API

---

Sample output

One question record looks like this:

{
  "questionId": 78394512,
  "title": "How do I chunk a PDF for RAG with LangChain?",
  "body": "I'm trying to build a RAG pipeline...",
  "link": "https://stackoverflow.com/questions/78394512/",
  "tags": ["python", "langchain", "vector-database", "rag"],
  "score": 2,
  "viewCount": 47,
  "answerCount": 0,
  "isAnswered": false,
  "hasAcceptedAnswer": false,
  "creationDate": "2026-04-19T14:22:00.000Z",
  "author": {
    "userId": 123456,
    "displayName": "devfounder99",
    "reputation": 347,
    "profileUrl": "https://stackoverflow.com/users/123456/"
  },
  "matchedKeywords": ["vector", "langchain"],
  "sourceKind": "tag",
  "sourceValue": "langchain"
}

Every field ready to drop into a CRM, a Slack channel, or a Notion database.

---

Pricing

First 50 questions per run are free. After that you pay per extracted question. No seat licenses, no tier gating. A 200 question run lands under $1 on the Apify free plan.

---

FAQ

Do I need a Stack Overflow account to use this scraper? No. The Stack Exchange API is fully public. You run it without signing in anywhere.

What is the Stack Exchange API rate limit? 300 calls per day anonymously. That covers hourly runs across 4 or 5 tags. For more, get a free key from stackapps.com/apps/oauth/register and the cap rises to 10,000 per day.

Can I scrape Stack Exchange sites other than Stack Overflow? Yes. Set site to any Stack Exchange slug: serverfault, superuser, dba, askubuntu, datascience, ai.stackexchange, devops.stackexchange, security.stackexchange.

How do I find a Stack Overflow tag slug? Open the tag page. The URL ends with the slug. Example: stackoverflow.com/questions/tagged/langchain means the slug is langchain.

Does the actor pull full question bodies? Yes by default. Turn off includeBody if you only need titles and metadata and want to save dataset space.

Can I filter for unanswered questions only? Yes. Set onlyUnanswered: true. For devtool lead gen this is the best filter since unanswered questions are live intent.

What is the difference between onlyUnanswered and onlyNoAcceptedAnswer? onlyUnanswered returns questions with zero answers. onlyNoAcceptedAnswer returns questions that may have answers but none were accepted. Both are strong lead gen signals.

Does it dedupe across runs? Yes. Question IDs are stored in the key value store under SEEN_IDS. Set dedupe: false to disable.

Can I run it on a schedule? Yes. Apify Scheduler goes down to 1 minute. Hourly covers most use cases.

Is scraping Stack Overflow allowed? This actor uses the official Stack Exchange API v2.3, which is rate limited and public by design. No HTML scraping, no terms of service risk.

---

Related Scrapemint actors

• Hacker News Scraper for stories and comments matching keywords
• Reddit Lead Monitor for subreddit and keyword mention tracking
• Upwork Opportunity Alert for freelance lead generation
• Product Hunt Launch Tracker for competitor launch monitoring
• App Store Review Scraper for mobile apps on iOS and Android
• Trustpilot Brand Reputation for DTC and ecommerce brands
• Google Reviews Intelligence for local businesses
• Indeed Company Review Intelligence for employer branding
• Amazon Review Intelligence for product review mining

Stack these to cover every public developer and customer conversation surface one brand touches.