🔍 Hacker News Scraper & Tech Trend Tracker

Extract stories and metadata from Hacker News' official Firebase API, with optional trend enrichment — engagement velocity, topic classification, domain breakdown — and structured JSON straight into an Apify dataset.

Pricing (live, as published on the Apify Store)

Event	Price
apify-actor-start	$0.00005 per event (one event per GB of run memory, minimum one)
Result — one story row	$0.05 per item

That is $50 per 1,000 stories, or 20 stories per $1.

Cost breakdown: 30 stories = $1.50. 100 stories = $5.00.

You are charged per story row delivered to your dataset:

• The trend summary in Tracker mode is written to the run's OUTPUT key-value record, not to the dataset, so it is not a billed result.
• Comments attached to a story ride inside that story's row and add no extra billable results.
• A run that delivers no stories — empty feed, blocked source, or no searchType chosen — charges no result events.

Input parameters

Parameter	Type	Range	Default	Description
searchType	string	top, new, best, ask, show, job	none — you must choose	Which HN feed to read. There is deliberately no default: an empty input never fetches or bills anything.
outputMode	string	tracker, raw	raw in code, tracker prefilled in the UI	tracker adds engagement/topic/domain fields to every row and writes a trend summary to OUTPUT. raw gives flat story records.
maxResults	number	1–100	30	How many stories to extract. Each delivered story is one billable result.
includeComments	boolean	true/false	false	Attach up to the top 10 top-level comments per story. This is not the full comment tree.

Output

One row per story. Field names as delivered:

{
  "id": 49570669,
  "title": "Actively exploited sandbox RCE in all Chromium versions",
  "url": "https://nvd.nist.gov/vuln/detail/cve-2026-85046",
  "hnUrl": "https://news.ycombinator.com/item?id=49570669",
  "author": "negura",
  "score": 386,
  "commentCount": 222,
  "type": "story",
  "time": 1788558721,
  "text": null,
  "topComments": [
    {"id": 49570901, "author": "tptacek", "text": "The sandbox escape is the interesting half.", "time": 1788559000}
  ]
}

topComments appears only when includeComments is true. In tracker mode each row additionally carries domain, pointsPerHour, commentsPerHour, discussionRatio, engagementTier, hoursOld, topics, age and contentType.

The Tracker summary (in OUTPUT, not the dataset) holds topicBreakdown, domainBreakdown, engagementTiers, contentTypeBreakdown, topAuthors, viralStories and generated insights.

Quick start

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("nexgendata/hacker-news-scraper").call(run_input={
    "searchType": "top",
    "outputMode": "tracker",
    "maxResults": 30,
    "includeComments": False,
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["score"], item["title"])

summary = client.key_value_store(run["defaultKeyValueStoreId"]).get_record("OUTPUT")
print(summary["value"]["insights"])

Real-world use cases

• Competitive intelligence — daily searchType: new, maxResults: 100, flag stories mentioning competitor names.
• Newsletter content — daily searchType: top, take titles plus top comments.
• Job board aggregation — searchType: job, extract companies and roles.
• Trend monitoring — outputMode: tracker, watch topicBreakdown and engagementTiers move week over week.

Run behaviour

• Genuine empty: if HN answers and the feed is empty, or every story in it has been deleted, the run succeeds with 0 rows and the status message says why. Nothing is charged for results.
• Source failure: a non-200, unreachable, or shape-changed response is retried (4 attempts on the feed, 3 per item) and then the run fails with a status message naming the cause. Nothing is charged for results.
• Time limit: the run derives a deadline from its own timeout, stops cleanly, says "Stopped early at the run time limit", and only the delivered rows are charged.

FAQ

Will this get blocked or rate-limited? It uses Hacker News' own public Firebase API, which is documented for automated access and needs no key. HN can still throttle or return errors under load; when that happens the run retries and then fails loudly rather than returning a silent empty result.

How fresh is the data? Live — the actor reads HN's current feed on each run.

Is my data private? Results stay in your Apify account.

How is this different from HN Algolia? Algolia is a full-text search index over HN history. This actor reads the live Firebase feeds (top/new/best/ask/show/ job) and enriches them; it is not a historical search.

Agentic payments

This actor supports autonomous payment via Skyfire, and is discoverable through Apify's MCP server (mcp.apify.com). The published pay-per-event pricing above applies unchanged.

Apify platform bonus

New to Apify? Sign up with the NexGenData referral link — you get platform credits on signup and help fund maintenance of this actor fleet.

---

A NexGenData utility actor.