Parasail Boasts Larger On-Demand GPU Fleet Than Oracle's Cloud

The cloud infrastructure scene is pretty much ruled by the big dogs: AWS, Microsoft's Azure, and Google Cloud. But when it comes to AI, the folks at Parasail think things are gonna shake out differently—and they're putting their money where their mouth is.

Parasail teams up with a bunch of providers to offer on-demand GPUs to companies and enterprises eager to build AI models and apps. They hook customers up with hardware like Nvidia's H100, H200, A100, and 4090 GPUs, and they claim it's way cheaper than what the big players charge.

"There's basically three cloud vendors who run the internet, but that's not how AI is shaping up," Tim Harris, one of Parasail's co-founders and the CEO of Swift Navigation, told TechCrunch. "It's way more spread out. The computing power is more flexible and fluid, so you can run it in a more horizontal way, and that's what we're all about."

Harris went on to say, "We didn't want a world where AI was controlled from start to finish by the hyperscalers."

Harris and Parasail CEO Mike Henry cooked up the idea for the startup a few years ago. Henry, who used to be the CPO at Groq, told TechCrunch he'd been mulling over what it would take to build AI infrastructure that could go toe-to-toe with Nvidia. He realized that with so many players jumping into AI infrastructure, there was a chance to make a horizontal move.

Henry pointed out that the fast pace of AI hardware innovation was making it tough for companies to keep up. "We had to really focus on making it as simple as possible for the customer," he said. "They're barely keeping up with the open source model releases alone."

Harris and Henry kicked off the company in 2023, brought on an engineering team, and started building in early 2024.

These days, there's no shortage of vendors eager to help enterprises and other companies build and scale their AI products. From big names like Nvidia and Microsoft to up-and-comers like Together AI and Lepton AI, customers have plenty of choices.

But the Parasail founders don't think these vendors are all cut from the same cloth. Sure, they all offer GPUs and AI infrastructure, Henry said, but he believes Parasail's proprietary tech sets them apart. This tech is what ties together Parasail's GPUs from different sources.

Wednesday marks the official launch of their platform, but Parasail is already working with dozens of customers, including Elicit, Weights & Biases, and Rasa. The company also raised $10 million from a seed round in 2024, with Basis Set Ventures, Threshold Ventures, Buckley Ventures, and Black Opal Ventures chipping in.

To really make a dent in the market, Parasail will have to go head-to-head with the hyperscalers and bet on the demand for GPUs continuing to grow. There's reason to believe it will, but there are also signs—like Microsoft canceling some data center contracts—that the predictions about needed AI infrastructure might be a bit over the top.

"We see literally no end to the demand," Harris said. "It's really that customers have a hard time doing it—have a hard time scaling AI. The models now are at a point where companies can just grab open source models and run them, but then getting access to GPUs, data centers, all the optimizations... we can do that with a click of a button."