Parse lets you build APIs for any website— 100x cheaper and faster than browser automation. Parse makes deterministic APIs that use direct network requests that cost fractions of a cent and return in milliseconds. ycombinator.com/launches/OnX-p… Congrats on the launch, @alexscraping!


genuine question how is this diff from firecrawl?


Stream, call, game, and work in even the most remote locations. Order online in minutes


Very expensive as compared to Firecrawler. Or, it's one can just use Crawl4ai.


would love to know the difference between this and @firecrawl_dev and @ExaAILabs


So slick! So I could take my voice agent, have setup an API with parse, then wrap that API with MCP to give the agent a new skill!?!?


Super cool, wish it worked


a YC company named Parse, eh. @ilyasu @jamesjyu


@alexscraping does it work for e-commerce sites with dynamic keywords and pagination?


Sounds cool! But how's the reliability compared to traditional methods? 🤔


Its an interesting idea - not sure how different from firecrawl - how it runs js which is main reason for headless - or how it's different from the api's gpts cursor can spawn with built in browser... but good luck!


i feel like this has been tried numerous times in the past, what makes this one different? AI?


Does it work on X? Asking for a friend.


does this work on sites that generally block these types of things? Like LinkedIn?


Pretty cool, but still requires traditional scraping for websites without private APIs?


How would it do if an internal site? Just VPN needed and it would work?


Very cool. Does this handle changes in the website’s layout etc?


This is sick Prices were higher than I expected but the determinism makes it worth it


Big one 🚀👀 Congratulations @alexscraping


How I remember the good ol’ days when parse was a backend app builder


finc gang


goated guy.


The best is that it's actually working


This is awesome


this is sooo cool


This is great; did you know Claude Code does this automatically? Would love to try this API. We did something internally that does the same, but I do think giving the code for scraping a site is necessary. The problems at scale of 100m requests are edge cases, captchas, honeypots


Indont get it, this tech already exists


YC funding software that crawls content you're not allowed to crawl. just like the AI companies, stealing.


I want to use this, excited


This is amazing!


Now this is going to solve a million problems out there.


Wow, this is quite something, Y! Building APIs this way sounds super efficient and a real game-changer, no?


It's make no sense unless I'm understanding something wrong


such a bold claim, cloudflare blocks them on scale, try the product on reddit


how do we bypass login in this approach ? tried scraping linkedin.com/jobs


United States Trends
Loading...

Something went wrong.


Something went wrong.