Why We Built Our Own Custom Chromium Stealth Browser

After 10+ years of helping customers collect public web data, we hit a wall: the browser. Here’s a behind-the-scenes look from our CEO, Neil Emeigh, on why we built our own.

When I started Rayobyte, I started it out of need:  there were no reliable, trustworthy, proxy providers on the market. I ran into it all:

• IPs were regularly offline
• Customer support would reply to me during outages and say: “We’re not too sure why they’re offline, if it doesn’t work for you then maybe you should look elsewhere.”
• ‘Dedicated IPs’ were double sold to more customers (I bought IPs from many sources and uncovered their unethical scheme through a confusing SQL query that said there were duplicates when I imported new IPs)

After chasing this around for a while, I had had enough and decided to create Blazing SEO (now Rayobyte) with a single target audience in mind:  people who needed proxies for SEO services.

Fast forward ten years, and we are now a top brand in the industry serving all ranges of data scraping products: data center, ISP, mobile proxies, and a scraping API.

And today, we add one more, critical, product to that offering: a stealth Chromium browser that can be self-hosted.

Early access: Stealth Chromium Browser

Please fill out this form here and we will be in touch..

Apply for Beta Access

The Journey To Becoming A Top Scraping Company

To be honest, we weren’t much of a scraping company up until the start of 2025. We had our product (scrapingrobot.com), but it wasn’t tended to or taken care of. Some of our headers being used for scraping were Chrome 120, while we’re on 143+ today. However, it worked for basic websites, while we focused our attention for the past years building the best-in-class proxy networks. 

Things changed when an opportunity arose at the end of 2024. A Fortune 200 ecommerce company approached us to scrape 200m+ pages/mo on a difficult Asian ecommerce website. For those who have scraped Asia ecommerce, you know the difficulties that were soon to be presented to us… 

Yet, we prevailed, and over a year later we’re their top vendor due to staying ahead of the target website’s constant, and aggressive, anti-scraping measures being put into place. 

But, this didn’t come easy. I suppose you might say we were “lucky” that this difficult website was the first serious scraping project we took on. It enabled us to start from core principles and build a solution that could be repeatable for all use cases. Today, we now scrape many websites at scale, including major SERP endpoints which have gotten increasingly difficult over the years too.

Introducing: [Stealth Browser]

I watched a Shark Tank episode 15 years ago and I’ve always remembered one thing Mark Cuban said: 

“In every entrepreneur’s journey, there is luck and timing”. You might say we were both, because the target website in December 2024 was surprisingly easy-ish to scrape. We didn’t need browsers, could use basic data center proxies, and so forth.

However, in Q1 2025, they ramped up. We had to go from no browsers and DC proxies, to generating high volumes of cookies and using residential proxies. We have the latter covered as a company, but the former… “What the heck are cookies? Why do we need them?!”. 

After understanding that we needed cookies to do lightweight HTTP scrapes, we came to the realization that we needed a good browser that could scrape at scale.

So, we went down the path of so many developers:  we first looked online for open-source projects that may be of help. Our criteria was simple, so we thought:

1. ✅ It should run on Linux servers (we use DigitalOcean, which has no GUI/Desktop interface, and no Windows support).
   
2. ✅ It should emulate unique fingerprints on every browser session. Different screen resolutions, different user-agents, different platforms, etc. We were operating at a large scale, and using the same fingerprint was sure to get banned quickly.
   
3. ✅ It should support Playwright (and not be detected when using Playwright). We weren’t interested in other automation frameworks that are not typical in web scraping.

What we found instead was:

1. ❌ There weren’t any open-source/free Chromium projects to build on. Chromium has 71% of the browser market share, and when scraping at millions of pages/day we had to ensure we blended in with real traffic, and other browsers (Firefox, etc) wouldn’t do.
   
2. ❌ The Chromium (paid) options that existed were built for Windows or desktop environments. I hopefully can safely speak for most developers: working on Windows at scale sucks 😢
   
3. ❌ The few Linux options we tried scaling with had downtime, lack of customer support, were detected by many websites, and broadly were not built for scrapers (instead, they’re built for the “account automation” industry).

And thus, our 1+ year journey in building a Linux-based stealth browser that can be self-hosted began!

Core Principles

The project has key core principles that we’ve built from the ground up:

1. It should work on Linux server environments that do not have GPUs
2. It should work with headful (via xvfb) and headless mode
3. It should emulate fingerprints from real-world devices around the world
4. It should be self-hostable for those who don’t want to pay for cloud hosting
5. It should be free to test and use for certain user segments
6. Lastly, but most importantly not the least, it should successfully bypass the major anti-bot systems as well as top scraping targets (across ecommerce, SERP, etc)

Early access: Stealth Chromium Browser

Please fill out this form here and we will be in touch..

Apply for Beta Access

Fingerprints

When using our stealth browser, you’ll be using real-world fingerprints collected from various websites we own. Our database contains thousands of permutations, ensuring that you can blend in with real-world traffic. 

Furthermore, you can customize your filters very granularly (want a Chrome 140 version, on Windows, which has screen resolution greater than 3000 pixels wide? No problem!). 

Bypasses major websites and anti-bot systems

We have a lot of gratitude for the inspiration that came from camoufox. Although it failed one of our tests (we needed Chromium, given its market share), we saw what was possible in starting our project.

Unfortunately, camoufox’s gift to the scraping world was, likely, one of the reasons for its demise:  because it was fully open-source, the major anti-bot engineers had an easy playbook to block it. We found that one anti-bot system even had a special tag for “camoufox” when it detected such requests. To fight such a battle, you need more engineering resources than your enemy, and camoufox, as far as we know, was only one guy up against the millions of dollars poured into the anti-bot companies.

And so, unfortunately, we won’t be open-sourcing our project for those watchful eyes. Furthermore, our license and TOS clearly state that usage from such companies is strictly prohibited.

With this in mind, we built our CI/CD pipeline to test many of the top websites that we currently serve with our scraping API. This gives you, the user, confidence that when updates are made from these target sites/systems, that you’ll have the commercial backing of Rayobyte to fight it off.

Self-hosted and free to use

Our license and usage model can be summarized as the following:

1. Free to use:
   1. pip install and you’re online. No registration required, fully self-hosted.
   2. Limitation: you’re only allowed to run one browser instance per host machine.
2. Free to use (unlimited threads):
   1. The same freedom as above, but you can run an unlimited number of browsers per machine.
   2. Limitation: to enable this plan, you need to be using a rotating residential, rotating data center, or rotating ISP proxy plan from Rayobyte when running the browser
3. Paid threads (no proxy limitation)
   1. If you want to Bring Your Own Proxy, and self-host with unlimited threads, then you can do so by purchasing concurrent browser threads.
   2. Pricing: please contact [email protected]
4. Cloud browser
   1. For those who don’t want to self-host, we will be introducing a cloud browsing solution in the near future. Please contact [email protected] to express interest to be one of the first to test.

Interested?

We’re now taking interested beta testers to try out the system. We’re using it at scale for our top scraping targets, but we recognize it may fail the test on other websites and so we need your help!

Requirements for testers:

1. Using Linux
2. Scraping at least 50,000 scrapes/mo

Early access: Stealth Chromium Browser

Please fill out this form here and we will be in touch..

Apply for Beta Access

Thank you, and happy scraping!

** NOTE: We only condone scraping on publicly available data that is not unethical, illegal or in any way a violation of any privacy. **