PRICING

Built around your work.
Priced around your compute.

Build in Model Studio with the compute that suits you.
Or work with our team on a model made for your business.

BRING YOUR COMPUTE

Your RunPod account

Use your own GPU connection in Model Studio.

Provider ratesBilled directly by RunPod

Connect your RunPod API key
Choose your GPU and storage
Train and use models in Studio

Open Model Studio

RunPod GPU and storage charges apply to your account.

MARC1 MANAGED CLOUD

Your model. Our setup.

Prepaid compute, managed from your workspace.

GPU cost + 20%Prepaid credits · No subscription required

Add credits when you need them
Review the GPU rate before starting
Set a budget for each allocation

Get started in Studio

GPU time is charged while allocated, including startup and idle time. Stop compute when you finish.

FINE TUNING AS A SERVICE

Built with your team

A specialized model for your enterprise workflows.

Tailored scopeLet’s define your project together

Data and use-case discovery
Model development and fine-tuning
Evaluation against your standards

Get in touch

Scope and cost depend on your model, data, compute, and deployment requirements.

A few things to know.

When does Managed Cloud billing start?

Billing starts when a GPU is allocated. Startup, model loading, training, and idle time all count. You review the hourly rate and budget before starting.

Can I use my own RunPod account?

Yes. Connect your RunPod API key in Model Studio and choose your GPU. RunPod bills your GPU and storage usage directly.

What happens when my budget runs out?

Managed Cloud stops accepting new work and shuts down the allocation as its budget is reached. Shared models need active compute and available budget to answer new requests.

Can Marc1 build a model for us?