Lock In Capacity. Guaranteed Availability. Volume Pricing.
Reserve GPUs for weeks, months, or longer. Get guaranteed capacity that won't disappear, volume pricing that beats on-demand rates, and a dedicated team that handles sourcing and setup.
Why Reserve with Spheron
Guaranteed Capacity
On-demand availability fluctuates. Reserved commitments don't. Lock in the exact GPU models and quantities you need for your project timeline.
Volume Pricing That Scales
The longer you commit and the more GPUs you reserve, the better your rate. For large deployments, we negotiate custom pricing with our data center partners.
Custom Cluster Configurations
Need 8 to 512+ GPUs with InfiniBand interconnect? Specific CPU, RAM, and storage requirements? We configure clusters to your exact specifications.
Dedicated Setup and Support
Submit your requirements, and our team handles procurement, configuration, networking, and ongoing support. You get a deployment-ready cluster.
How It Works
From request to deployment in four steps. Our team handles the complexity so you can focus on your models.
Tell Us What You Need
Fill out a quick form with your GPU model, quantity, duration, interconnect preferences, and target start date.
We Source the Best Option
Our team checks availability across certified data center partners, negotiates pricing, and sends you a quote. Typical turnaround: 24 to 48 hours.
Review and Confirm
Review the proposal covering hardware specs, pricing, location, and terms. Once confirmed, we begin provisioning.
Deploy and Scale
Your reserved hardware is configured and ready. Access your cluster, run your workloads, and scale up by adding more capacity at any time.
Need More Than What's Listed?
Reserved Capacity
Commit to a duration, lock in availability and better rates
Custom Clusters
8 to 512+ GPUs, specific hardware, InfiniBand configs on request
Supplier Matchmaking
Spheron sources from its certified data center network, negotiates pricing, handles setup
Tell us your GPU needs and we'll match you with the right provider from our certified data center network.
Typical turnaround: 24–48 hours
What You Can Reserve
One Quote. Transparent Pricing.
Tell us what you need and we handle the rest. We source hardware across multiple data center partners, negotiate on your behalf, and deliver a single, competitive quote. No back-and-forth with multiple vendors. No surprises on your invoice.
Compare Instance Types
On-Demand
Full-price, guaranteed availability. Deploy instantly, pay by the minute.
Production workloads, long-running training, critical inference
Spot
Discounted pricing using idle capacity. Same hardware, up to 50% off.
Batch jobs, experiments, hyperparameter tuning, flexible workloads
Reserved
CurrentLocked capacity with volume pricing. Custom clusters and dedicated support.
Large deployments, multi-month projects, enterprise teams
Reserved Commitment FAQ
What is a reserved GPU commitment?
A reserved commitment guarantees you a specific number of GPUs for a set duration at a negotiated price. Unlike on-demand, your capacity is locked in and allocated to you. No availability surprises, and volume pricing that improves with scale and duration.
How long can I reserve GPUs for?
Reservations start at 1 week and extend to 12+ months. Longer commitments typically unlock better pricing. Contact our team to discuss your timeline and we'll find the right term.
Contact our team →What happens after I submit a reservation request?
Our team reviews your requirements, checks availability across our data center partner network, and returns a quote within 24 to 48 hours. The quote covers hardware specs, pricing, location, and terms. No obligation until you confirm.
Can I change my reservation after it starts?
Yes. If your needs change, we can work with you to adjust GPU quantities, extend duration, or modify your cluster configuration. We're flexible because we know project requirements evolve.
Is there a minimum GPU count for reserved commitments?
Reserved commitments are designed for larger deployments, typically starting at 8 GPUs. For smaller, shorter-term needs, on-demand instances with per-minute billing may be a better fit.
Learn about on-demand instances →Secure Your GPU Capacity
Tell us your GPU requirements, timeline, and budget. Our team will source the best option and have a quote back to you within 48 hours.