Hacker News Scraper — Who Is Hiring Jobs + HN Search
Pricing
from $1.60 / 1,000 job results
Turn the monthly Ask HN: Who is hiring? thread into structured job JSON, or full-text search ALL of Hacker News by keyword. Optional AI enrichment (BYO key) extracts company, role, salary, stack, remote & visa; AI trend digest summarizes the whole thread. Delta mode for alerts.
Which of the three monthly HN "whoishiring" threads to scrape. hiring = "Ask HN: Who is hiring?" (companies posting jobs, the default). seeking = "Ask HN: Who wants to be hired?" (people looking for work). freelancer = "Ask HN: Freelancer? Seeking freelancer?".
Optional. When set, the Actor IGNORES the monthly thread and instead full-text searches ALL of Hacker News (every story and comment) for this phrase via the official Algolia HN Search API, newest first — e.g. rust remote, your product name, a competitor. Combine with “Max age (hours)” to bound recency and “Only new since last run” for mention alerts. Up to 1000 newest matches per run (the API's ceiling). Leave empty for the classic Who-is-hiring mode.
What to search when “Search query” is set: all (stories + comments, default), stories only, or comments only.
Optional case-insensitive substring match on the full comment text (e.g. remote, rust, senior). Non-matching comments are dropped before billing, so you only pay for matches. Leave empty to return every comment in the thread.
Maximum number of job postings to return. Set 0 for no limit (all postings in the thread). Applied before AI enrichment, so you never pay to enrich rows you capped out.
Only return postings newer than this many hours. Set 0 to disable age filtering and return all postings in the thread regardless of age.
Delta / alert mode: only output comments not seen on a previous run that also had this flag on. Already-seen comments are dropped before push (not billed), so this is the cheapest way to run this Actor on a cron schedule and only pay for genuinely new postings. State is tracked per Actor in a dedicated key-value store, keyed by HN comment id. See README “Delta mode / alerts”.
Adds aiCompany / aiRole / aiLocation / aiSalary / aiTechStack / aiRemote / aiVisa / aiEmploymentType to every posting via the Anthropic, Mistral or OpenAI API (pick which below). Unlike regex-based scrapers that guess from the pipe convention, the model returns null / “unknown” / empty whenever the prose doesn’t clearly state a value — it never fabricates. Requires your own API key for the provider you pick — billed separately by that provider, not by this Actor (see README “AI enrichment”). If turned on without a matching key, enrichment is skipped (with a warning) and postings are still returned normally, without the ai* fields.
Adds ONE extra summary row (id trend-digest, never billed) produced by a single LLM call across every posting in the run: prose summary, top technologies with mention counts, salary observations, remote/hybrid/onsite split, top locations and notable trends. Great for a monthly hiring-market report from the Who-is-hiring thread. Needs the same BYO API key as AI enrichment (works with either add-on independently). The digest only reports what the postings literally say — counts and figures are never invented.
Which AI provider runs enrichment (only relevant when “AI enrichment” is on). anthropic (default) uses Claude via anthropicApiKey. mistral uses a Mistral model via mistralApiKey instead — pick this if you’d rather bring a Mistral key than an Anthropic one. openai uses a GPT model via openaiApiKey.
Your Anthropic API key (sk-ant-…). Only used when “AI enrichment” is on and aiProvider is anthropic; billed separately by Anthropic. Not required unless aiEnrichment is on.
Claude model used for AI enrichment when aiProvider is anthropic (only relevant when “AI enrichment” is on). Haiku is fast and inexpensive; Sonnet gives higher-quality extraction on longer, more nuanced comments.
Your Mistral API key. Only used when “AI enrichment” is on and aiProvider is mistral; billed separately by Mistral. Not required unless aiEnrichment is on with aiProvider=mistral.
Mistral model used for AI enrichment when aiProvider is mistral (only relevant when “AI enrichment” is on). Small is the default — it matches larger Mistral models on this well-scoped extraction task at a fraction of the cost.
Your OpenAI API key (sk-…). Only used when “AI enrichment” is on and aiProvider is openai; billed separately by OpenAI. Not required unless aiEnrichment is on with aiProvider=openai.
OpenAI model used for AI enrichment when aiProvider is openai (only relevant when “AI enrichment” is on). Defaults to gpt-4.1-mini — cheap, fast and ample for this extraction task. Any chat-completions model works, including the gpt-5 family.
