Serpex.dev
Enrich your AI & Apps with structured, real-time search data
83 followers
Enrich your AI & Apps with structured, real-time search data
83 followers
Serpex.dev offers a REST API/SDKs for real-time, structured Search data, giving LLMs and SaaS apps live search insights without scraping complexities. Great for AI applications seeking current search intelligence and data projects.
Hello Product Hunt 👋
I’m Kartey, solo founder & founding engineer behind this project.
I want to share the story of how this product came to life from a real problem I faced while building.
🧠 It all started while building my Research AI Agent
While working on a research-focused AI agent, I realized something crucial: AI is powerful but without real-time information, it’s basically working with outdated knowledge.
About 80%+ of my queries needed live web data to stay accurate and relevant. Most existing APIs couldn’t deliver Reliable, fresh information, though some sources are costly and geared toward enterprise use.
That’s when I started digging deeper into the problem.
🚧 The Problem
Every available search API I tested came with major trade-offs:
💸 Too expensive for indie founders or small startups
⚙️ Too complex to integrate or maintain
🚫 Easily blocked or rate-limited by engines.
It was frustrating, I just wanted something simple, affordable, and reliable.
💡 The Realization
At some point, it clicked, instead of building another research tool dependent on someone else’s API, why not build the foundation itself?
So I pivoted and started building a Unified Search API focused on what truly matters:
✅ Fast & accurate: responses under 3-4 seconds, 5k+ requests per minute (benchmarked up to 50k+ per second)
✅ Easy to integrate: Typescript & Python SDKs ready
✅ Affordable: starting at just $0.0008/request
✅ Multi-engine support: Supporting Google, Bing, DuckDuckGo, Brave. Adding more.
✅ Reliable & structured: clean data, no captcha, no random blocks.
✅ Auto handle: routes your queries across best engines and retries.
⚙️ Who It’s For
🧑💻 Solo founders & indie hackers building AI products
🧠 Researchers & data enthusiasts exploring trends and insights
⚡ No-code / low-code builders integrating live data into workflows
🚀 Developers & AI engineers who want up-to-date info without the API headache
We support Google, Bing, DuckDuckGo and Brave. Adding more soon.
No complex setup. No overpriced plans. No random blocks. Just clean, real-time search results ready for your AI and data-driven projects.
❤️ Why I Built It
it was something I needed while building my own project. But as I refined it, I realized many other builders faced the same pain so now it’s not just a personal tool, it’s a way to help developers, founders, and researchers unlock the freedom to access live data easily.
💬 Let’s Talk
If you were building your own AI agent today… What kind of real-time data would you want it to access first?
Would love your thoughts, feedback, or even crazy use cases you imagine.
Thanks for reading and for supporting me ❤️
@kartey I’ve dealt with the same pain while building AI workflows, most search APIs either break under volume or return messy HTML that needs tons of post-processing.
Curious how do you handle consistency across engines? For example, if Google throttles or changes SERP markup, do you have fallback strategies or a normalization layer in place?
Love the pricing and SDK-first approach, feels built by a developer, for developers.
@freyaisnice Thanks Freya! We support Google, Bing, DuckDuckGo, Yandex, Yahoo, and Brave. For reliability, we use fallback engines and thousands of non-blocking proxies to prevent tracking.
As the developer, I built this for both developers and no-code platforms like n8n and Make.com. We provide complete documentation, API support, and a typed API for easy debugging and use.
I love that you're not asking for subscriptions, just allowing to purchase affordable credit packs!
@guillaume_duvernay Yes, because I dislike subscriptions, I prefer one-time purchases. Use whatever you want; your credits won't expire, and you only pay for what you use.
@kartey I'm the same, trying to embed that no subscription approach into Lookio that I'm building now :)
I tried the tool out and was surprised with the order of some results - can I ask which location is being used? Is there any way we can pick the location (e.g. UK google results)?
@guillaume_duvernay Yes, we previously had that feature, but I removed it because Google was providing inconsistent results. Since we use thousands of proxies, Google can't track accurately. Using a US proxy to query Google UK sometimes returned irrelevant results. Google uses the `cc` parameter in the search URL for country-specific results, which gives us 95% success. I'm working on advanced API customization, but for now, I've removed the feature because it wasn't working as expected. We should have it back in the next release in 1-2 weeks.
For quicker responses, join my Discord server where I'm more active and readily available for help and support, I will do my best to add any features the customer needs.
@kartey thanks for the response!
Brilliant if it works. I'll definitely check it out to see if it helps with our more difficult scraping projects. And love the pay-as-you go pricing.
@tomsi thanks Tom, my primary focus was the pricing with quality and I am working hard to delivering it.
If you need any help or support, just join our discord server, i'm active on it, and will get faster response or support.