product-hunt-scraper-api

A Product Hunt scraper for Python, built on ScrapingBee's Product Hunt scraper API. It returns the ranked leaderboard as objects with rank, name, slug and URL.

Verified against the live site on 2026-08-25: the homepage returned 70 ranked products and Auto-Mode charged 1 credit for the call.

Install

pip install product-hunt-scraper-api

Today's launches

from product_hunt_scraper_api import ProductHuntScraper

scraper = ProductHuntScraper("YOUR-API-KEY")

for product in scraper.leaderboard()[:10]:
    print(product.rank, product.name, product.url)

1 akta.pro https://www.producthunt.com/products/akta-pro
2 Diet Claude https://www.producthunt.com/products/diet-claude
3 Jotform AI Data Assistant https://www.producthunt.com/products/jotform

Key from app.scrapingbee.com, 1,000 credits free.

Why this costs 1 credit and not 25

Product Hunt public listing pages render their product links server-side. That means no headless browser is needed, and the cheapest proxy tier is enough.

Rather than guessing that, the client sends mode=auto. ScrapingBee then tries configurations from cheapest to most expensive and charges only for the one that returns the page. On the homepage that settled at 1 credit. Confirm it yourself:

products, charged = scraper.leaderboard_with_cost()
print(len(products), "products for", charged, "credits")

If a page ever hardens, Auto-Mode climbs on its own up to max_cost without a code change. Raise the ceiling for a stubborn page:

scraper.leaderboard(max_cost=75)

What a Product object holds

Product Hunt uses Tailwind utility classes with no stable card identifier, so the dependable anchor is the product link, a[href^="/products/"]. Its text carries the rank inline, as "1. akta.pro". The client splits that apart for you:

product.as_dict() returns all four, ready for a dataframe or a JSON column.

Tracking a topic

for product in scraper.topic("developer-tools")[:10]:
    print(product.rank, product.name)

Any Product Hunt listing URL works through leaderboard(url=...), so category, topic and date-archive pages all behave the same way.

Rolling your own parse

When the four fields are not enough, take the HTML and parse it yourself:

html = scraper.page_html("https://www.producthunt.com/products/akta-pro")

render_js=True is available but rarely needed on listing pages, and it raises the floor from 1 credit to 5.

Watching launch day

Product Hunt rankings move all day. A poll every 30 minutes at 1 credit is 48 credits a day, which makes a full launch-day watch effectively free:

import time

seen = {}
for _ in range(12):
    for product in scraper.leaderboard():
        previous = seen.get(product.slug)
        if previous and previous != product.rank:
            print(f"{product.name}: #{previous} -> #{product.rank}")
        seen[product.slug] = product.rank
    time.sleep(1800)

Check the balance first with scraper.usage(), which is free and capped at 6 calls a minute.

Cost

Tiers at scrapingbee.com/pricing.

Related scrapers

For adjacent launch and startup research the same key reaches Google search, Google News and AI extraction, which is handy when a page changes shape and you would rather describe fields than maintain selectors.

Scope

Public, pre-login pages only. Scraping behind login credentials is prohibited by the ScrapingBee terms. Keep API keys out of AI coding assistants.

MIT licensed. Repository . extraction rules