Product Hunt Scraper: Daily, Weekly & Monthly Launches, Topics, Makers, Websites

Pull Product Hunt launches as clean JSON: name, tagline, description, the product's real website, votes, comments, reviews, daily/weekly/monthly/yearly rank, launch date, topics, makers, media and optional top comments. Ask by launch day, week, month or year, by topic, by product or launch URL, or by free text.

What you can do with it

• Daily launch feed: schedule yesterday every morning and send the top products to Slack, a newsletter or a database.
• Weekly and monthly leaderboards: last week, 2026-W41, last month, 2026-10, or paste a leaderboard URL.
• Competitor and market research: follow every launch in a topic (topic:artificial-intelligence) or a theme (crm, open source llm) and see what gets votes.
• Lead generation: each product's own website and domain (not a Product Hunt redirect), plus makers' names, usernames, headlines, Twitter handles and profile links.
• Trend and sentiment analysis: turn on top comments to see what early users say.

Input

Field	What it does
queries	One per line, mix freely (see the table below).
maxPerQuery	Most products any one query returns (default 100).
maxItems	Cap for the whole run and your cost. Each query gets a fair share; what one query doesn't use goes to the next.
order	Automatic (leaderboard rank for one day, most votes for longer windows, newest for the rest), or Leaderboard rank, Most votes, Newest, Featured time.
featuredOnly	Only launches Product Hunt featured.
topic, postedAfter / postedBefore	Restrict latest, topic and text queries to a topic or a date window.
includeComments, commentsPerProduct	Add each product's most upvoted comments (default 10).
keywords	Keep only products that mention one of these words. Filtered products are not charged.
maxWaitMinutes	Longest total pause the run may take when Product Hunt asks it to slow down (see Limits).

Query forms

You type	You get
today, yesterday, 2026-10-14	That launch day, in Pacific time like Product Hunt's own leaderboard.
this week, last week, 2026-W41	That leaderboard week (Monday to Sunday).
this month, last month, 2026-10, year:2026	That month or year.
2026-10-01..2026-10-07, 2026-W40..2026-W41	Any range, both ends included.
latest	The newest launches.
topic:developer-tools or a /topics/... URL	Launches in that topic. A /categories/... URL is read as the topic of the same name.
slug:notion, a /posts/... or /products/x/launches/... URL	That one launch.
a /products/raycast URL	The launches listed on that product's page.
a daily, weekly, monthly or yearly /leaderboard/... URL	That leaderboard's featured launches (add /all for every launch).
anything else, e.g. crm	Launches whose name, tagline, description or topics contain every word, newest first.

Example:

{ "queries": ["yesterday", "last week", "topic:developer-tools", "crm"], "maxPerQuery": 50, "maxItems": 200 }

Output

One item per product:

{
  "id": "1100001", "name": "Beam.win", "slug": "beam-win", "tagline": "Do the thing faster", "description": "...",
  "url": "https://www.producthunt.com/products/beam-win",
  "website": "https://beam.win", "websiteDomain": "beam.win", "websiteRedirect": "https://www.producthunt.com/r/XR2QV6TXWMZHHX",
  "votesCount": 412, "commentsCount": 37, "reviewsCount": 3, "reviewsRating": 5,
  "dailyRank": 1, "weeklyRank": 4, "monthlyRank": 12, "yearlyRank": null,
  "launchDate": "2026-10-19T07:01:00Z", "featuredAt": "2026-10-19T07:01:00Z", "featured": true,
  "thumbnail": "https://ph-files.imgix.net/...", "media": [{ "type": "image", "url": "https://ph-files.imgix.net/...", "videoUrl": null }],
  "topics": ["Productivity", "Developer Tools"], "topicSlugs": ["productivity", "developer-tools"],
  "makers": [{ "name": "Jane Doe", "username": "janedoe", "headline": "Founder", "twitter": "jane", "website": null,
               "profileUrl": "https://www.producthunt.com/@janedoe" }],
  "makersHidden": 1, "hunter": "someone", "comments": [], "query": "yesterday"
}

• website is the product's own site with Product Hunt's tracking removed; websiteRedirect keeps the Product Hunt link.
• Product Hunt hides some makers' identities. Those are never returned as placeholder names: makersHidden counts them.
• A product that two of your queries both find is returned (and charged) once.

The run's SUMMARY record (in the key-value store) lists every query: what it is, how many launches the window holds, how many were returned, and its status (ok, no results, failed, or not reached when a cap stopped the run first).

Pricing

Pay per result: $1.20 per 1,000 products on the Free plan, down to $0.90 per 1,000 on paid Apify plans (product-scraped). Websites, ranks, media and comments are included. Keywords, duplicates and caps apply before you are charged. A run that finds nothing at all fails instead of silently returning an empty dataset; a query with no matches is simply reported as no results.

Limits and notes

• Speed: Product Hunt serves about 1,200 products per 15 minutes (about 600 with 10 comments each; measured Oct 2026). Runs within that finish in seconds to a few minutes; a larger run pauses until Product Hunt allows more, and logs how long it waits. maxWaitMinutes (default 30) caps the total pause; when it's reached the run stops with what it has and the summary says so. For a whole month (25,000+ launches), raise maxWaitMinutes and the run timeout.
• Text search finds launches in the best-matching topic (crm -> the CRM topic, newest first), or among the most recent launches when no topic matches. Use a topic, slug: or a URL when you know exactly what you want.
• Not included: promoted (paid) placements never appear in results. Product Hunt's newer category pages are read as the topic with the same name. The product websites themselves are not visited (no emails).
• Data is public Product Hunt launch data, subject to Product Hunt's terms.

Feedback and requests

We welcome feedback: a field or option you wish this had, a change to how it works, or a scraper you would like us to build. Open an issue in the Issues tab of this page and tell us what you need. We read every one, and we build the requests that come up most. If this actor works well for you, a short review helps other people find it.