Hacker News Search — Stories, Comments & Live Front Page
Pricing
from $0.35 / 1,000 result items
Search all of Hacker News (stories, comments, Ask/Show/Launch HN, polls, jobs) by keyword with points, comments, date, author, domain and story filters, or pull the live top/new/best/ask/show/jobs lists. Optional nested comment trees and author profiles.
Keywords to search. Algolia syntax: wrap a phrase in double quotes for exact match ("vector database"), prefix a word with - to exclude it (python -django). Leave empty to list items by date (kind=story → newest stories matching the filters).
Additional search terms processed one after another in the same run — e.g. several brand names or competitors to monitor. Each output item carries the query that found it. maxItems is a total across all queries.
Which HN item type / list to pull. story, comment, ask_hn, show_hn, launch_hn, poll, job, front_page, newest, any search the Algolia index and support every filter below. top, new, best, ask, show, jobs are the live ranked lists from the official Firebase API (same order as the HN pages) — query, author, storyId, dates and searchIn are ignored for them.
relevance and date are native Algolia sort orders (relevance falls back to date when there is no query). points and comments are applied client-side to the fetched result set (up to maxItems, max 1000 per query) — combine with dateFrom/sinceDays or minPoints to make the set meaningful. For live Firebase lists the default is the list's own ranking.
Restricts which attributes Algolia matches the query against (restrictSearchableAttributes). all also matches story text, URLs and comment text — right for brand/mention monitoring, noisy for topic lists: use title for "stories about X". url is handy for finding every submission of a site: query="github.com/apify", searchIn="url".
Let Algolia match misspellings (one typo per word). Off by default: with it on, "LLM" also matches "Limbo", "Limits" and "GLM", "Kafka" matches "Kakapos". Turn on to catch misspelled mentions.
Keep items with at least this many points. Comments have no points on HN — ignored for kind=comment.
Keep stories with at least this many comments (Algolia num_comments). Ignored for kind=comment.
Only items created in the last N days. 0 = all time. Ignored when dateFrom is set.
Earliest creation date, YYYY-MM-DD or ISO 8601 (UTC). Overrides sinceDays.
Latest creation date (inclusive; a bare date means end of that day UTC), YYYY-MM-DD or ISO 8601.
HN item id, e.g. 4058874. With kind=comment returns the comments of that thread (searchable by query). Alternative to includeComments when you want flat, paginated comments of a single story.
Keep only items whose URL is on this domain or its subdomains, e.g. apify.com or arxiv.org. If query is empty the domain is also used as the Algolia URL query, so listing all submissions of a site is one call.
Total items to output across all queries. Algolia serves at most 1000 hits per query; live lists hold 200–500 ids. (Legacy name limit is still accepted.)
Deprecated alias of maxItems, kept for backward compatibility. maxItems wins when both are set.
Attach the discussion of each story as a nested comments tree (author, date, plain text, children). Costs one extra request per story with comments; use maxCommentsPerStory/commentsDepth to bound the payload.
Total comments kept per story when includeComments is on (breadth-first: top-level replies first, then their children).
Attach authorProfile (karma, account creation date, about, number of submissions) from the official HN API. One request per distinct author (cached within the run).
Truncation length (characters) for text and comment texts after HTML is converted to plain text.
Output each external URL only once (first hit wins; scheme, www. and trailing slash are ignored). Useful when the same article was submitted several times.
Keep only these fields in each item (id is always kept). Leave empty for the full item. Known fields: id, type, kind, query, title, url, domain, hnUrl, points, numComments, author, createdAt, text, storyId, storyTitle, storyUrl, parentId, tags, comments, authorProfile, source, sourceUrl, fetchedAt.
Comment tree depth
commentsDepth
Optional
How many reply levels to keep in
comments(1 = only direct replies to the story).