Hacker News Scraper - Stories, Comments, Users & Who Is Hiring
Pricing
from $0.14 / 1,000 items
Hacker News Scraper - Stories, Comments, Users & Who Is Hiring
Scrape Hacker News: top / new / best / Ask / Show / job lists, a day's top stories, search with date, points and comment filters, full comment trees, user profiles, Who is hiring and Who wants to be hired posts parsed into rows, keyword trends. Monitor mode. $0.14 per 1,000 items.
Pricing
from $0.14 / 1,000 items
Rating
0.0
(0)
Developer
Top Api
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
7 days ago
Last modified
Categories
Features
Share
What does Hacker News Scraper do?
Hacker News Scraper extracts data from Hacker News through its two public APIs (the official Firebase API and HN Search by Algolia) and returns one clean row per story, comment, poll, job ad, user profile, hiring post or trend.
Hacker News API · HN scraper · Hacker News comments scraper · Who is hiring scraper · Hacker News search export
- Every list HN ranks: top (the front page), new, best, Ask HN, Show HN and jobs, with each story's rank, read live from the official API.
- Search with real filters: keywords, item type (stories, comments, Ask HN, Show HN, jobs, polls, on the front page now), date window, minimum points and comments, authors, domains, title-only or link-only matching. Filters run inside HN Search's own query, so a filter that keeps 1% of the matches reads 1% of the pages.
- No 1,000-result cap: HN Search answers 1,000 hits per query; the actor walks back through time and keeps going (5,000 rows in one run measured, more is fine).
- Whole discussions: every comment of a story in HN's own ranked order, with depth, parent, story title and replies, as rows or nested inside the story.
- User profiles: karma, about, account age, submissions, story and comment counts, last activity, plus their stories and comments.
- Monthly hiring threads parsed into rows: Who is hiring? (company, roles, locations, remote / onsite / hybrid, regions, salary, equity, visa, stack, emails, apply links, hiring system, YC batch), Who wants to be hired? (location, remote, relocation, technologies, email, résumé, GitHub, LinkedIn) and Freelancer? Seeking freelancer?
- Trends: a keyword's stories and comments in a window against the window before (growth, points, top stories, top domains, trend score), and the words and phrases rising fastest in story titles.
- Monitor mode: on a schedule, a run delivers only what earlier runs did not, optionally pushed to Slack, Discord or a webhook. A poll with nothing new costs about $0.00008.
- No login, cookies or API key. Pay only for delivered rows; filtered rows are never charged.
What data can you get from Hacker News?
| 📰 Title, link, domain, text | 🔺 Points, comment count, rank | 👤 Author, author karma and account age |
| 💬 Comments with depth, parent, story, rank | ⚡ Points per hour, comments per hour, isTrending | 🗳️ Poll options with votes |
| 🧑💻 User karma, about, created, story / comment counts, last active | 💼 Job posts: company, roles, salary, remote, visa, stack, apply links | 🙋 Candidates: location, remote, relocation, technologies, email, résumé |
| 📈 Keyword trends with growth and top stories | 🚀 Rising terms in titles | 🔗 Every link inside a text |
How to scrape Hacker News with Hacker News Scraper
- Open the actor and pick what to read: a list (prefilled: Top), search queries, HN links, item ids, usernames, a day, hiring threads or trend keywords. Mix them freely in one run.
- Optionally set filters (type, dates, points, comments, authors, domains) and extras (comments, author profiles, user submissions).
- Set Max rows per source (prefilled 30) and click Start. Most runs finish in 1-5 seconds.
- Download the rows as JSON, CSV, Excel or HTML, or read them through the API, a webhook or MCP.
⬇️ Input
| Field | Type | What it does | Default |
|---|---|---|---|
lists | array | top, new, best, ask, show, job | top when nothing else is asked |
searchQueries | array | One HN Search per line; "exact phrase" in quotes | - |
startUrls | array | item?id=, user?id=, submitted?id=, threads?id=, from?site=, news / newest / best / ask / show / jobs, front?day= (see Limits), hn.algolia.com search links | - |
itemIds | array | Story, comment, poll or job ids, e.g. 8863 | - |
usernames | array | Profile rows, e.g. pg | - |
topStoriesDays | array | 2026-11-01, today, yesterday: that day's stories ranked by points | - |
hiringThreads | array | whoIsHiring, whoWantsToBeHired, freelancer | - |
hiringMonths | array | latest, 2026-10, October 2026, last 3 | latest |
trendKeywords | array | One trend row per keyword | - |
discoverRisingTerms | boolean | Rising words and phrases in titles | false |
itemType | string | story, comment, ask_hn, show_hn, job, poll, front_page, all | story |
sortBy | string | relevance or date (newest first) | relevance |
searchIn | string | all, title, url, text, author | all |
exactMatch | boolean | No typo tolerance ("rust" will not match "trust") | false |
dateFrom / dateTo | string | 2026-11-01, ISO, unix time, or "7 days" | - |
postedWithinHours | integer | Only the last N hours | - |
minPoints / minComments | integer | Stories with at least that many | - |
authors / domains | array | Only these usernames / sites; alone they list everything they match | - |
filterKeywords | array | Title or text must contain one of these (useful on lists) | - |
includeComments | boolean | Each story's discussion | false |
maxCommentsPerStory | integer | 0 = whole discussion | 100 |
maxCommentDepth | integer | 1 = top level only, 0 = every level | 0 |
commentSort | string | ranked (HN's order), newest, oldest | ranked |
commentsFormat | string | rows (one per comment) or nested (inside the story) | rows |
includeAuthorProfile | boolean | Author karma, created, account age, submissions on every row (free) | false |
includeUserSubmissions / submissionTypes / maxSubmissionsPerUser | A user's stories and/or comments, newest first | false / story / per source | |
hiringKeywords, hiringKeywordsMatch, technologies, locations, remoteOnly, onsiteOnly, visaOnly, withSalaryOnly, withEmailOnly, willingToRelocateOnly, latamOnly | Hiring-row filters | off | |
trendWindowHours | integer | Window length; compared with the window before | 24 |
risingTermsMax, risingMinMentions, risingMinGrowthPercent | integer | Rising-term limits | 25, 3, 100 |
monitorMode, monitorName, resetMonitor | Only rows new since the last run, one memory per name | off | |
webhookUrl, slackWebhookUrl, discordWebhookUrl, alertMinPoints | Push each run's new rows | - | |
maxResultsPerSource | integer | Per list, search, day, user or hiring month | 100 (a hiring month: every post) |
maxResults | integer | Cap for the whole run, comments included; 0 = none | 0 |
compactOutput | boolean | Drop HTML, link lists and id lists | false |
includeDead, maxConcurrency, proxyConfiguration | Advanced | false, 32, no proxy |
Input names used by other Hacker News actors are accepted too (mode, query, story_ids, username, max_results, include_comments, comment_limit, comment_depth, maxItems, tags, since, hiringMonth, trend_queries, hours, detectTrends, onlyNew ...), so an API call can switch without a rewrite.
Input JSON
{"searchQueries": ["\"vector database\"", "rust"],"itemType": "story","postedWithinHours": 168,"minPoints": 50,"includeComments": true,"maxCommentsPerStory": 50,"maxResultsPerSource": 200}
{"hiringThreads": ["whoIsHiring"],"hiringMonths": ["last 3"],"remoteOnly": true,"technologies": ["Rust", "Go"]}
Put filters in the query
Type, author, date window, points, comments and "search in" go into HN Search's own query, so they cost nothing extra. On lists (top, new ...) the filters are checked on each item read: the new list with minPoints: 50 read all 500 items and kept 23 in 1.8 s.
⬆️ Output
Every row has a recordType: item (story, comment, poll, poll option, job), user, job-post, candidate or trend, and a source saying which input produced it. The dataset has six views: Stories, Comments, Users, Who is hiring, Who wants to be hired, Trends.
A story from the top list (real run, 2026-11-07):
{"recordType": "item", "id": 50399657, "type": "story", "storyType": "story","title": "CATL Sodium Battery Hits LFP Cost Parity in 2026: 15,000 Cycles, 30-Year Life","url": "https://ievchina.com/energy-storage/catl-sodium-battery-cost-parity-lfp-2026-15000-cycles/","domain": "ievchina.com", "author": "doener", "points": 86, "numComments": 34,"createdAt": "2026-11-06T23:00:46.000Z", "hnUrl": "https://news.ycombinator.com/item?id=50399657","rank": 1, "kidsCount": 7, "ageHours": 1.4, "pointsPerHour": 60.26, "commentsPerHour": 23.82, "isTrending": true,"listName": "top", "source": "list:top"}
A comment on it (same run):
{"recordType": "item", "id": 50400039, "type": "comment", "author": "EwanToo","text": "Alternative site which is still up\n\nhttps://www.pv-magazine.com/2026/11/06/sodium-ion-batteries-approach-cost-parity-with-lfp-with-70-gwh-covered-by-supply-agreements/","createdAt": "2026-11-06T23:42:41.000Z", "parentId": 50399657, "storyId": 50399657, "depth": 0, "rank": 1,"storyTitle": "CATL Sodium Battery Hits LFP Cost Parity in 2026: 15,000 Cycles, 30-Year Life","links": ["https://www.pv-magazine.com/2026/11/06/sodium-ion-batteries-approach-cost-parity-with-lfp-with-70-gwh-covered-by-supply-agreements/"]}
A user profile:
{"recordType": "user", "id": "pg", "karma": 157316, "about": "Bug fixer.", "createdAt": "2006-10-09T18:21:32.000Z","accountAgeDays": 7333, "submittedCount": 15565, "storyCount": 699, "commentCount": 9985,"lastActiveAt": "2024-03-11T12:57:11.000Z", "lastStoryAt": "2023-08-27T00:51:13.000Z", "hnUrl": "https://news.ycombinator.com/user?id=pg"}
A Who is hiring? post (fields shown are a subset; text and textHtml carry the whole post):
{"recordType": "job-post", "id": 50338732, "threadTitle": "Ask HN: Who is hiring? (November 2026)", "month": "2026-11","headerLine": "Simple Analytics | DevOps Engineer | REMOTE (worldwide, any timezone) | Full-time, 32h, 24h possible | €5,000/month (full-time) | https://www.simpleanalytics.com/jobs/devops","company": "Simple Analytics", "domain": "simpleanalytics.com", "fundingStage": "Bootstrapped","roles": ["DevOps Engineer"], "employmentTypes": ["full-time"], "locationPolicy": "remote", "remoteRegions": ["Worldwide"],"salaryMin": 5000, "salaryMax": 5000, "salaryCurrency": "EUR", "salaryPeriod": "month","stack": ["Elasticsearch", "ClickHouse", "Ansible", "OpenTofu", "Docker", "Grafana", "GitHub Actions", "TypeScript"],"applyUrls": ["https://www.simpleanalytics.com/jobs/devops"]}
A Who wants to be hired? post (subset):
{"recordType": "candidate", "id": 50338729, "hnUser": "LoganDark", "threadMonth": "2026-11","location": "US", "countries": ["United States"], "remote": true, "willingToRelocate": false, "yearsExperience": 10,"email": "logandark@logandark.net", "githubUrl": "https://github.com/LoganDark", "hasContact": true, "contactScore": 2}
A keyword trend (72-hour window, subset):
{"recordType": "trend", "kind": "keyword", "term": "AI", "windowHours": 72,"storyCount": 631, "previousStoryCount": 544, "storyGrowthPercent": 16, "commentCount": 3036, "previousCommentCount": 3013,"totalPoints": 10631, "avgPoints": 16.8, "maxPoints": 1292, "uniqueAuthors": 515, "trendScore": 88, "trendDirection": "steady","topDomains": [{ "domain": "github.com", "count": 54 }, { "domain": "youtube.com", "count": 17 }],"topStories": [{ "id": 50364336, "title": "AI is forcing us to pay off 40 years of technical debt", "points": 1292, "numComments": 748 }]}
How complete the parsed hiring fields are
Hiring posts are free text; each parsed field is a reading of the post and the full text always comes with it. Share of rows with the field filled, measured on 431 Who is hiring? posts (November 2026 plus the three months to November) and 520 candidate posts:
| Job posts | filled | Candidates | filled |
|---|---|---|---|
| company, headerLine | 100% | location | 99% |
| locationPolicy (remote / onsite / hybrid) | 95% | remote | 95% |
| domain | 86% | technologies | 92% |
| roles | 77% | 88% | |
| employmentTypes | 75% | résumé link | 75% |
| stack | 67% | years of experience | 52% |
| applyUrls | 60% | website | 41% |
| remoteRegions | 48% | GitHub | 39% |
| contactEmails | 38% | 32% | |
| salary (min, max, currency, period) | 35% | time zones | 33% |
| equity | 24% | ||
| hiring system (Greenhouse, Lever, Ashby ...) | 21% |
Visa (5%), relocation (6%), funding stage (14%) and YC batch (2%) are filled only when a post says so.
How can I use Hacker News data?
- Brand and competitor monitoring: schedule
searchQuerieswith monitor mode and a Slack webhook; every new story or comment that names you arrives once. - Launch tracking:
itemType: show_hn,postedWithinHours: 168,minPoints: 100(17 Show HN stories in the week measured). - Recruiting and job-market research: the monthly hiring threads as a table of companies, roles, salaries and stacks, or of candidates with contact links.
- Trend research: keyword trends against the previous window, and rising terms.
- Datasets for analysis or AI: whole discussions with structure (depth, parent, rank), user histories, past days' top stories.
How much does it cost to scrape Hacker News?
Pay per event, no start fee:
| Event | Price | Per 1,000 |
|---|---|---|
| Item (story, comment, poll, poll option, job; a nested comment counts as one) | $0.00014 | $0.14 |
| User profile | $0.0009 | $0.90 |
| Job post (Who is hiring / SEEKING FREELANCER) | $0.0009 | $0.90 |
| Candidate post (Who wants to be hired / SEEKING WORK) | $0.0036 | $3.60 |
| Trend row (keyword or rising term) | $0.0045 | $4.50 |
Measured runs: the top 30 = $0.0042; a 1,000-row search = $0.14; two stories with their 1,675 comments nested inside (1,677 items) = $0.23; November 2026's Who is hiring? (205 posts) = $0.18 and Who wants to be hired? (504 posts) = $1.81. Author profiles, poll options and the dataset views are free. Rows removed by filters, users or items that do not exist, and rows a monitor delivered before are never charged. Set a maximum charge per run and the actor stops cleanly at it (a $0.003 limit on the top list returned 21 rows).
Run it on a schedule
- Save your input as a task, turn on Only new since the last run and give the watch a Monitor name.
- Add a schedule (every hour, every day). Each run delivers only rows earlier runs did not; searches read newest first.
- Optionally add a Slack, Discord or webhook URL to get each run's new rows pushed, and
alertMinPointsto push only popular ones. - Or call it from the Apify API, Make, Zapier, n8n or an AI agent over MCP (
https://mcp.apify.com/?actors=TopApi/hacker-news-scraper).
A monitor poll looks at the same rows a normal run would (the first N of each list or search); rows delivered before keep their place and are skipped. Measured: a poll with nothing new cost $0.000076-0.000078; a poll that found 3 new stories cost $0.0001.
Coverage
- Lists: top 500, new 500, best 200, Ask HN, Show HN and jobs as HN publishes them (41, 186 and 31 on 2026-11-06).
- Search: every story, comment, poll and job in HN Search's index (stories since 2006); 1,000 hits per read, deeper by walking back in time.
- Comments: the whole tree of any story in one read (two discussions with 1,675 comments and author profiles in 7.4 s).
- Hiring threads: every month since the threads began; the Freelancer thread is found by its title.
Limits & empty runs
- HN's own past front page (
front?day=) is not read: news.ycombinator.com answers this actor's client with HTTP 419 on every route. Afront?day=link orfrontPageDatesreturns that day's stories ranked by points instead (topStoriesDays), which is not the same list: 15 of the 30 stories on HN's page for 2026-11-01 were on it. - Points are published for stories and polls only; comments have none.
- HN Search's points and comment counts trail the live site by a point or so; lists and item ids are read live.
- A user's
storyCount/commentCountcome from HN Search;submittedCountis HN's own total (it also counts deleted items). - The largest discussions with Author profile on each row mean one profile read per commenter: HN's three biggest threads (12,286 comments) took 249 s at the default 256 MB (634 s at 128 MB). Give such runs 1,024 MB to finish sooner; the price is the same.
- An input that matches nothing ends in about a second with a message saying so, and charges nothing.
FAQ
Does Hacker News have a public API?
Yes: the official Firebase API (lists, items, users) and HN Search by Algolia (full-text search). This actor reads both and adds filters, depth past the 1,000-hit cap, parsing of the hiring threads, trends and monitoring.
Can I get the comments of a story?
Turn on Comments with item ids, links, lists or searches. Comments come in HN's ranked order (rank on top-level comments), with depth and parentId; commentsFormat: nested puts them inside the story row instead.
Can I get more than 1,000 search results?
Yes. Set maxResultsPerSource above 1,000; past the first 1,000 the rows are read newest first (5,000 rows of "google" in 7-50 s, $0.70 at the item price).
Why did my run return no rows?
The status message says which source matched nothing and which filters removed rows, with counts. Loosen the filters or widen the date window.
Do I need a Hacker News account or API key?
No.
Your feedback
Found a field you need or a post the hiring parser reads wrong? Open an issue on the actor's page.