Vetter VM - early access
Cloud servers, sized to the job.
Pick a machine by vCPU, memory and NVMe storage, from small shared instances to GPU and single-tenant bare metal. We are opening access in small groups.
- Type
- High Performance
- Plan
- hp-4
- vCPU
- 4 dedicated
- Memory
- 16 GB
- Storage
- 200 GB NVMe
- Network
- IPv4 + IPv6
Taken from the High Performance table below.
Four kinds of machine
Start small and move to a bigger class when the workload needs it.
-
Cloud Compute
Shared vCPU instances for websites, small APIs, bots and dev boxes.
View plans - Cloud Compute -
High Performance
Dedicated vCPU on current-generation CPUs, for databases and busy services.
View plans - High Performance -
Cloud GPU
GPU instances for model inference, fine-tuning and batch jobs.
View plans - Cloud GPU -
Bare Metal
A whole single-tenant server with no hypervisor between you and the hardware.
View plans - Bare Metal
Every plan includes
NVMe storage
Local NVMe disks on every instance class.
Root access
Full control over SSH, with your own keys.
IPv4 and IPv6
A public address of each kind per instance.
Bandwidth included
A monthly transfer allowance listed with each plan.
Plans
These are the planned early-access specs and may change before launch. Prices come at launch; ask by email and we will send them to you first.
Shared vCPU.
| Plan | vCPU | Memory | NVMe | Bandwidth | Price |
|---|---|---|---|---|---|
| vc-1 | 1 | 1 GB | 25 GB | 1 TB | Early access - Request access for vc-1 |
| vc-2 | 1 | 2 GB | 50 GB | 2 TB | Early access - Request access for vc-2 |
| vc-4 | 2 | 4 GB | 80 GB | 3 TB | Early access - Request access for vc-4 |
| vc-8 | 4 | 8 GB | 160 GB | 4 TB | Early access - Request access for vc-8 |
| vc-16 | 8 | 16 GB | 320 GB | 5 TB | Early access - Request access for vc-16 |
Dedicated vCPU.
| Plan | vCPU | Memory | NVMe | Bandwidth | Price |
|---|---|---|---|---|---|
| hp-2 | 2 | 8 GB | 100 GB | 4 TB | Early access - Request access for hp-2 |
| hp-4 | 4 | 16 GB | 200 GB | 5 TB | Early access - Request access for hp-4 |
| hp-8 | 8 | 32 GB | 400 GB | 6 TB | Early access - Request access for hp-8 |
| hp-16 | 16 | 64 GB | 800 GB | 8 TB | Early access - Request access for hp-16 |
One or more GPUs per instance.
| Plan | GPU | vCPU | Memory | NVMe | Bandwidth | Price |
|---|---|---|---|---|---|---|
| gpu-24 | 1 x 24 GB | 8 | 32 GB | 250 GB | 5 TB | Early access - Request access for gpu-24 |
| gpu-48 | 1 x 48 GB | 12 | 64 GB | 500 GB | 8 TB | Early access - Request access for gpu-48 |
| gpu-80 | 1 x 80 GB | 16 | 128 GB | 1 TB | 10 TB | Early access - Request access for gpu-80 |
| gpu-80x4 | 4 x 80 GB | 64 | 512 GB | 4 TB | 20 TB | Early access - Request access for gpu-80x4 |
Physical cores, single tenant.
| Plan | vCPU | Memory | NVMe | Bandwidth | Price |
|---|---|---|---|---|---|
| bm-16 | 16 cores | 64 GB | 2 x 1 TB | 10 TB | Early access - Request access for bm-16 |
| bm-32 | 32 cores | 128 GB | 2 x 2 TB | 15 TB | Early access - Request access for bm-32 |
| bm-64 | 64 cores | 256 GB | 4 x 2 TB | 20 TB | Early access - Request access for bm-64 |
How early access works
- 1
Email us
Tell us what you want to run and which plan looks right.
- 2
Get a slot
We reply with an access date and the price for your plan.
- 3
Deploy
You get your machine and a direct line to the team running it.
Questions
Can I deploy a server today?
Not yet. Vetter VM is in early access, and we add people by email in small groups.
How much does it cost?
Prices are not public yet. Email us with the plan you want and we will send the price with your access date.
Where are the servers?
We will publish regions before launch. If you need a specific location, say so in your email.
Who runs Vetter VM?
Vetter AI. The same team builds the product and answers your email.
Want a machine when we open?
One email gets you on the list.