A few months ago, I was developing a bookmark manager for <a href="https://fika.bar" rel="nofollow">https://fika.bar</a>. I needed a way to fetch metadata, screenshots, and markdown content from websites. Since Fika is a donation-based product, I couldn't afford any of the providers out there, so I decided to build instead of buy.<p>"Easy," I thought. But weeks later, I found myself building data pipelines, implementing proxy rotation schemes, and dealing with TLS fingerprinting. Not fun—and definitely not how I wanted to spend my time. My product was a bookmark manager, not a data extraction service, yet most of my efforts ended up there!<p>It turns out that the Internet has become a difficult place to extract information from. Many websites have implemented measures against data extraction, making it more challenging and expensive than it should be to build products that rely on web data.<p>Handinger is my attempt to fix this. During that time, I also noticed that people training LLMs need to fetch data at scale. So, I extracted the work I'd done for Fika and built a very affordable, general-purpose scraping API. I managed to optimize it so I can price it at only 0.0005 EUR per URL—peanuts!<p>Try it out and let me know what you think :)
Hi , if you want to check if our sdk fits your need for extraction, we can discuss any solution since "I couldn't afford any of the providers out there" . <a href="https://github.com/docwire/docwire">https://github.com/docwire/docwire</a>
Hi, Congrats on launching the product.<p>Curious to know how you arrived at this pricing which looks really good. I have experienced that a larger cost in systems like these are the data transfer costs. Would be great if you could share some insights on that too.