Hacker News Trends and Stories (author-free)
Pricing
Pay per usage
Hacker News Trends and Stories (author-free)
Hacker News front-page, new, best, Ask, Show and job stories, keyword search, and keyword/domain trend summaries from the official HN API and the HN Search API. Author handles and text are never included.
Pricing
Pay per usage
Rating
0.0
(0)
Developer
ORLIBACHAR Ltd
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
6 days ago
Last modified
Categories
Features
Share
Get Hacker News stories as clean rows, or summarise what HN is talking about by keyword or by linked domain.
- Story lists: top, new, best, Ask HN, Show HN and jobs, in HN's own ranking order.
- Keyword search: stories matching a query, newest first, with date and score filters.
- Trends: one row per keyword or per domain with story count, total points, total comments, average points and the highest-scoring stories.
The Actor reads two public APIs that need no key and no login: the official Hacker News API and the HN Search API. It does not use a browser or a proxy and runs in 256 MB.
This Actor is independent. It is not made, endorsed or supported by Y Combinator, Hacker News or Algolia.
What it does not return
No author usernames, no story or comment text, no comment threads and no user profiles. Every output row is built from a fixed list of keys.
Story titles and URLs are returned as published. A title can itself mention a person.
Input
| Field | Type | Meaning |
|---|---|---|
mode | text | top, new, best, ask, show, job, search or trends. Default top. |
query | text | Required for search. |
trendBy | text | domain (default) or keyword. Trends only. |
keywords | list | One trend row per keyword (trendBy = keyword). |
dateFrom / dateTo | YYYY-MM-DD | Story date window in UTC. dateTo is inclusive. Trends default to the last 7 days. |
minScore | number | Keep stories with at least this many points. |
maxItems | number | Stop after this many rows. Default 50. |
includeCommentCounts | yes/no | Fill comment counts. Default yes. |
topStoriesPerTrend | number | Stories listed inside each trend row. Default 3. |
maxStoriesScanned | number | Stories read per keyword (or in total for domain trends). Default 1000. |
Examples:
{ "mode": "top", "maxItems": 30 }
{ "mode": "search", "query": "rust", "minScore": 50, "dateFrom": "2026-10-15", "maxItems": 100 }
{ "mode": "trends", "trendBy": "keyword", "keywords": ["rust", "golang", "zig"], "dateFrom": "2026-11-01", "dateTo": "2026-11-13" }
How the modes work
- List modes read HN's current list (up to 500 ids for top/new, up to 200 for ask/show/job) and fetch each story. Date and score filters are applied to the fetched stories, so a strict filter may return fewer rows than
maxItems. Deleted and dead items are skipped. Jobs usually have no score, so they drop out whenminScoreis above 0. - Search asks the HN Search API for stories, newest first. Date and score filters are applied by the API.
- Trends read up to
maxStoriesScannedstories (newest first) in the window and add them up. If the limit is hit, the row saystruncated: trueand only covers the newest stories in the window. Domain trends put text posts without a link under(text post, no link).
Output
Story rows (list and search modes):
| Key | Meaning |
|---|---|
id | HN item id |
url, domain | Linked URL and its host without www. (empty for text posts) |
title | Story title |
type | story, job or poll |
score | Points |
commentCount | Number of comments |
time | Posted at, UTC, ISO 8601 |
hnUrl | Link to the HN discussion |
rank | Position in the HN list, or in the search results |
Trend rows (trends mode):
| Key | Meaning |
|---|---|
trendType, key | keyword or domain, and the keyword or domain |
storyCount, totalPoints, totalComments, avgPoints | Totals over the stories read |
topStories | Highest-scoring stories: id, title, url, score, commentCount, hnUrl |
window | The from / to dates used |
minScore | Score filter used |
approxMatches | Keyword trends: match count reported by the search API (may be approximate) |
truncated | true if maxStoriesScanned was reached |
rank | Position in this run's output |
Pricing
Pay per row written to the dataset: one price for story rows, a higher one for trend rows. See PRICING.md (not final until a measured test run).
Limits
- Scores and comment counts change over time. Each row is a snapshot at run time.
- Keyword matching is done by the HN Search API. It can match words in the URL or other indexed text, not only the title.
- Search pages move backward in time, so they are not limited to Algolia's page depth. If more than 100 stories share the exact same second, some of them can be missed.
- List modes make one request per story. Pauses: 0.05 s between stories, 0.4 s between search pages. Rate-limit and server errors are retried with backoff.
Data sources
- Official Hacker News API: https://github.com/HackerNews/API
- HN Search API: https://hn.algolia.com/api