For most of my eighteen years in SEO (and now AIO, GEO or whatever you want to call it), web crawlers have handed me the same choice: run out of memory, or pay a small fortune.
Memory or money
Crawling a large site on my laptop often turned into an exercise in memory management. Moving to the cloud fixed that, but the cost jumped.
I grumbled about that gap for years. Then, last year, I took it to David Mortiboy (now my Co-Founder at Scuba) and we decided to do something about it.
So we set out to build the crawler I'd wanted for years.
Building for efficiency and scale
Our brief was simple: a cloud crawler that scales smoothly without the usual cloud price tag. David has spent nearly ten years building software for well-known companies in the UK and US. Together we designed around efficiency from the first line of code, because that was the only way to make cloud crawling affordable.
We also wanted Scuba to work for marketers and developers alike, handling modern SEO and GEO tasks as well as the essentials. That meant a mix of expected functionality and modern improvements from day one, including:
Control and scale
Concurrent crawls, scheduled runs, and side-by-side comparison views to track what changed between releases.
Data extraction
Custom extraction by CSS selector, XPath, or regex, with structured CSV exports.
Team collaboration
Multi-seat access, shared crawl views, inline annotations, and colleague tagging on key fixes.
AI-native connectivity
Native MCP server support, enabling AI assistants to run crawls and query data directly.
AI-native, not AI bolted on
Most tools in this market were built for a human clicking through a dashboard, with AI added years later on top. We built Scuba the other way round, with an MCP server (the standard that lets AI assistants work directly with your tools) at its core.
An assistant can start a crawl, inspect the results and answer questions about a site without anyone exporting files in between. Ask it what broke on a site this week, and the answer comes straight from the crawl data.
Pricing
Gone are the days of carving a big site into several local crawls and merging the exports afterwards, and no high prices for cloud crawling. There is a free plan for small sites, and Lite is £45 a month for 50,000 URLs. Full details are on our pricing page.
Looking ahead
Launching is just the start. What we build next depends entirely on the people using it. Community feedback will drive our product roadmap, meaning you will have a direct say in what we build moving forward. Whether that is shaping upcoming features like autonomous monitoring and deeper integrations, or pitching entirely new ideas, you can help steer what comes next.
Scuba launches on 9 December 2026. If you are looking for a cost-effective way to crawl sites at scale and query data natively with AI, join the waitlist and we will email you on the day.
Ashley Williams
Co-Founder, Scuba