Artificial intelligence infrastructure startup Parasail Inc. today announced the launch of its new AI Deployment Network, a new cloud-hosted platform that provides enterprises with immediate, cost-efficient and on-demand access to powerful graphics processing units.

Parasail says its AI Deployment Network is special because it allows companies to tap into a massive, contract-free pool of high-performance GPUs at highly competitive rates, making it easy to scale any AI workload in a matter of minutes. It provides access to a wide range of GPUs, including Nvidia Corp.’s beefy H100 and H200 chips, plus A100 and 4090 processors for more specialized tasks.

The startup further claims that its on-demand AI compute platform provides the most optimal performance of any GPU resource provider too, thanks to its proprietary orchestration engine that matches workloads with its global network of GPUs. Through this, it says, it’s able to solve cost permutation problems for customers while eliminating any complexities around workload management.

In other words, Parasail simplifies access to AI infrastructure, providing computing resources that can be accessed at any time of day or night and with no commitments, to support rapid deployments, AI experimentation and scale existing production workloads.

The startup was founded by AI infrastructure veterans Mike Henry and Tim Harris and is backed by $10 million in seed funding from investors including Basis Set Ventures, Threshold Ventures, Buckley Ventures and Black Opal Ventures.

Henry, the startup’s chief executive, said there’s a lot of demand for true “on-demand access” to GPU resources, because few infrastructure providers offer it at the same scale as his company does.

“In reality, legacy cloud providers use small amounts of compute capacity to lure customers into long-term contracts,” Henry said. “At Parasail, we are providing the first real-time, true on-demand access to massive compute without the hidden constraints.”

With Parasail, companies can deploy production-ready endpoints and gain access to dozens of GPUs within a matter of hours, with minimal setup or configuration involved, allowing them to accelerate almost any kind of AI workload, the startup claims. In doing this, they can benefit from a two-to-five times cost advantage versus other infrastructure providers, it says.

The startup gathered a slew of testimonials from early adopters to back up its claims. For instance, Elicit Research PBC, which has built an AI-powered research assistant similar to Microsoft Corp.’s Researcher agents and Google LLC’s Deep Research tool, tapped Parasail’s cloud to screen hundreds of thousands of scientific papers on demand.

“Elicit is using LLMs to screen more than 100,000 scientific papers each day, but the cost of high-quality, real-time processing was prohibitive,” said Elicit CEO Andreas Stuhlmüller. “Parasail was essential for removing this bottleneck.”

The popular LLM development platform Weights & Biases Inc. is another customer, using Parasail’s cloud to support its deployment of the popular DeepSeek R1 reasoning model when it took the industry by storm earlier this year. “Parasail moved at lightning speed to get us set up with massive DeepSeek capacity and top-shelf throughput,” said Weights & Biases Chief Technology Officer Shawn Lewis. “They will give you the latest and greatest faster than anyone else.”

Parasail has certainly made a splash and there’s a lot to like about the on-demand GPU access it provides. That will be all the more true if it can cater to the large-scale deployment requirements of the world’s biggest enterprises in future. That said, it’s far from alone in offering simpler access to GPU resources these days.

One of its biggest rivals looks to be CoreWeave Inc., which has built a GPU cloud that’s dedicated to AI workloads and notably went public last week through a somewhat disappointing initial public offering. Other rivals include GMI Cloud Inc., which is focused on the rapid deployment of GPUs and claims to be able to do so in seconds. There’s also Together Computer Inc., which has raised millions of dollars in capital from backers including Salesforce Inc. And the decentralized cloud storage company Storj Labs Inc. got into the GPU game last year after acquiring a startup called Valdi Labs PBC.

Though the industry is competitive, Henry said the prospects are looking good for Parasail because most enterprises are shifting toward a multi-supplier aggregation infrastructure model.

“The future of AI infrastructure isn’t about a single cloud provider,” he said. “It’s about an interconnected network of high-performance compute providers.”

Image: SiliconANGLE/Microsoft Designer