Today, we’re unveiling two major additions to our cloud platform: On-Demand GPUs and Reserved GPUs. 

Here’s what’s new:

  • High reliability: 99.5% uptime guarantee on all instances

  • Premium hardware: Initially focused on H100 and H200 GPU configurations

  • Streamlined experience: Simple selection process for GPU model, storage, and image

  • Quick provisioning: Deploy instances within minutes

  • Tiered support: Tier 1 support with response times under 24 hours

  • Flexible pricing: On-demand hourly rates and discounted long-term options

Check out the latest demo created by our Product Lead, Kai Huang, now live on X and Youtube. 

As our CEO and Co-founder Dr. Jasper Zhang discussed in a recent letter, Hyperbolic is uniquely positioned to launch these product lines. We've earned the trust of over 195,000 developers who use our platform for fine-tuning, training, and experimenting with the latest models.

Over the past year, we've built a platform that's easy to use, affordable, and purpose-built for AI/ML workloads. We've vetted data centers globally, saving our users countless hours and resources. We've proven our expertise in compute infrastructure through our Spot GPUs and serverless inference products by serving our user base and partnering with leading teams at Hugging Face, Vercel, Quora, Coinbase, Reve AI, Stanford, NYU, UC Berkeley, UC San Diego and more. Now, we're bringing the same advantages that made us the choice for individual developers to solve the biggest challenges facing business and enterprise builders.

Our new cloud offerings are curated product lines that connect customers with high-performance, reliable GPU instances. Unlike our Spot GPU model, which utilizes excess capacity from suppliers, our On-Demand GPU and Reserved GPU leverage suppliers' best machines, maintained to strict reliability standards.

This is the first look at our two new product lines, we'd love your feedback to help us improve before the official launch.

You can get started by signing into your Hyperbolic account at app.hyperbolic.ai and sharing your feedback via email at support@hyperbolic.ai.

About Hyperbolic

Hyperbolic is the on-demand AI cloud made for developers. We provide fast, affordable access to compute, inference, and AI services. Over 195,000 developers use Hyperbolic to train, fine-tune, and deploy models at scale.

Our platform has quickly become a favorite among AI researchers, including those like Andrej Karpathy. We collaborate with teams at Hugging Face, Vercel, Quora, Chatbot Arena, LMSYS, OpenRouter, Black Forest Labs, Stanford, Berkeley, and beyond.

Founded by AI researchers from UC Berkeley and the University of Washington, Hyperbolic is built for the next wave of AI innovation—open, accessible, and developer-first.

Website | X | Discord | LinkedIn | YouTube | GitHub | Documentation