When should I pick VPS over VDS or bare-metal?

Pick VPS when your workload spends most of its time idle waiting for network events. Trading bots, webhook handlers, light indexers, build runners, monitoring agents: anything that bursts CPU briefly and then waits. Every tier — VPS included — is validator-adjacent in Frankfurt on a private VLAN to the NLN fleet, so the differentiator isn't network access: pick VDS when you need pinned cores and RAM that never share a hyperthread with another tenant, with steal time at 0.00% by contract. Pick bare-metal when you're running a Solana validator, an MEV pipeline, or a Geyser plugin that needs 128+ GB RAM and multiple NVMe drives with no virtualization overhead.

How much does a Solana VPS cost?

vps.starter is $100/month (4 vCPU, 16 GB, 160 GB NVMe), vps.pro is $190 (8 vCPU, 32 GB, 320 GB NVMe), vps.enterprise is $350 (16 vCPU, 64 GB, 640 GB NVMe). Every tier ships on a 1 Gbps port with 2 TB of egress included. The price includes Solana-ready images, full root, DDoS-filtered transit, and the same dashboard and support queue as the rest of the NLN stack, plus an in-catalog upgrade to pinned-core VDS and validator-grade bare metal when you outgrow shared vCPU. Hosting and Solana data products on one bill.

Can I run a Solana validator on a VPS?

No. Mainnet validators need 256+ GB RAM, two NVMe drives in production, and 10 Gbps unshared. That's a bare-metal workload. A VPS is fine for a testnet validator if you want to learn the operator workflow: vps.enterprise (16 vCPU, 64 GB) can handle that. For mainnet, see bare-metal.

Which regions can I deploy a VPS in?

Frankfurt today. We can quote bespoke regions for committed annual contracts if you need other locations.

Do I get root access? Can I install whatever I want?

Yes, full root over SSH. We don't lock the kernel, we don't restrict outbound ports, we don't inspect what you run. The only enforcement is abuse rate limits on outbound SMTP and a fair-share network scheduler so one tenant can't saturate the uplink. Run Ubuntu, Debian, Arch, NixOS, or upload your own ISO.

How fast does the VPS provision?

Same-day for the standard Ubuntu 24.04 image; most orders go out within the hour during EU business hours; off-hours sit at the front of the queue and ship at the next opening. Custom images, larger storage volumes, or non-default network setups add a few hours. We don't over-provision the hypervisor pool, so there is real spare capacity sitting hot; you don't wait for a host to be carved.

What latency should I expect from my VPS to your RPC nodes?

Under 0.5 ms to every NLN Solana service. Every VPS is validator-adjacent in Frankfurt on a dedicated private VLAN to the NLN RPC, gRPC, and validator fleet, so your path never leaves our switch fabric. Don't take our word for it: on first boot, ping rpc.nln.clr3.org and time a real getSlot call; whatever you measure is your number, and it's a five-second test.

Shared Compute · VPSFrankfurt · validator-adjacent

Solana VPS in Frankfurt, sized for trading workloads

KVM servers in Frankfurt, validator-adjacent and private-networked to the NLN RPC and gRPC fleet — <0.5 ms to every NLN Solana service. Sized for chain-adjacent workloads: trading bots, webhook handlers, indexers. NVMe storage, generous egress, full root with your choice of OS, and a straight upgrade path to pinned-core VDS. From $100/mo.

Configure a server Talk to an engineer

vps.pro · first five minutes

$ ssh root@185.x.x.x
Welcome to Ubuntu 24.04 LTS · vps.pro · fra
$ apt install -y nodejs npm && npm i @solana/web3.js
added 42 packages in 6s
$ systemctl enable --now sniper-bot.service
Created symlink → /etc/systemd/system/sniper-bot.service
$ journalctl -u sniper-bot -f
slot 367201442 · processed
slot 367201443 · processed
# order to running bot, same day

Provisioning

Same-day

most within the hour, EU business hours

Access

Full root

KVM · any OS

NLN fleet path

<0.5 ms

private VLAN to RPC/gRPC

Uptime SLO

99.99%

monthly · automatic credits

from $100/mo

Catalog · 3 tiers

Configure a VPS

Every spec and price below is exactly what hits checkout. The catalog is generated from the same table that does the billing.

Need a shape that isn't on the rail: more drives, GPUs, 100 Gbps? We quote custom builds in 24 h.

vps.pro

The production default for trading bots that wait on the network.

$190/mo

monthly · cancel anytime

Compute

8 vCPU · shared EPYC

8 vCPU · fair-share scheduled on shared EPYC

Memory

32 GB DDR4 ECC

dedicated allocation, shared channels

Storage

320 GB NVMe

▮ 320 GB NVMe

local NVMe (host mdraid1) · nightly snapshots · restore < 15 min

Network

1 Gbps fair-share2 TB included egress

Contract

·Own kernel, own IP, full root
·Fair-share vCPU and network scheduling
·Local NVMe storage, nightly snapshots
·Key-only SSH from first boot, any OS
·Provisioning: Same-day · most within the hour (EU business hours)
·SLO: 99.99% monthly · automatic credits

Sized for

Production trading botsLight indexersSniper bots (bursty)

Not for: Hot decode loops at full gRPC volume: that is vds.standard territory.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee

What this is

A working box for chain-adjacent workloads

A Solana VPS is a regular KVM virtual server bought for one reason: running chain-adjacent workloads (trading bots, webhook handlers, indexers) without babysitting hardware. Ours ship from Frankfurt, validator-adjacent and on a dedicated private VLAN to the NLN RPC, gRPC, and validator fleet — under 0.5 ms to every NLN Solana service — with KVM virtualization, shared vCPU, NVMe storage, 1 Gbps network, and 2 TB of included egress on every tier. Three tiers: vps.starter at $100/mo (4 vCPU, 16 GB, 160 GB NVMe), vps.pro at $190/mo (8 vCPU, 32 GB, 320 GB NVMe), and vps.enterprise at $350/mo (16 vCPU, 64 GB, 640 GB NVMe). These are shared-CPU instances: the hypervisor fairly schedules vCPU time across tenants, which is fine for bursty workloads but means you should not run a mainnet validator or a 24/7 hot loop here. For that, see VDS or bare-metal, where the cores are pinned and steal time is 0.00% by contract.

What you are buying here is a working box, honestly described: KVM with full root, NVMe storage, 2 TB of included egress on every tier, and sizes shaped around how bots and indexers actually consume resources: short CPU bursts between long network waits. No inflated promises stapled to a commodity VM.

And because the workload rarely stays the same size, the exit is part of the product: when monitoring says you've outgrown shared vCPU, the same dashboard moves you to a pinned-core VDS, where the cores are yours alone and steal time is 0.00% by contract.

Tenancy, stated honestly

Shared vCPU, fairly scheduled, with a one-liner to measure steal time yourself and a clear trigger for when to move up.

Measure, don’t trust

We don’t publish latency numbers for VPS. Your first-boot checklist benchmarks the box in five minutes; your numbers are the real ones.

One vendor for the stack

The RPC, gRPC, and WebSocket fleet your bot consumes is ours too. Hosting and data on one bill, with an upgrade path to VDS and metal.

The fine print, in large print

What shared vCPU means, and when to leave it

CPU steal time

real, measure it with mpstat

Cache & memory bus

shared with neighbors

Failure blast radius

host fails → snapshot restore < 15 min

You pay for

a working box, not exclusivity

yours time-sliced other tenants host reserve

Shared vCPU is the part most hosting pages mumble through, so here it is plainly: your cores are scheduled time on silicon other tenants also use. If a neighbor pegs eight threads, you can lose cycles to context switching and cache pollution. For a bot that idles between slots, this is invisible. For a hot decode loop, it shows up exactly where it hurts: the 99th percentile.

The signal to watch is CPU steal time. Under 1% during peak Solana hours (13:00–17:00 UTC), you are fine on VPS. Consistently above 2–3%, the honest fix is not a bigger VPS. It is VDS, where the cores are pinned and steal time is 0.00% by contract.

measure it yourself · run during peak hours

# Steal-time check: 5 samples, all cores
mpstat -P ALL 1 5 | awk '/Average/ \
  {print "CPU " $2 " steal=" $6}'

# healthy VPS:   steal < 2%
# move to VDS:   steal > 3% sustained

We put the upgrade trigger in the docs on purpose. Mis-sold workloads churn; measured ones stay.

Sizing, consultatively

Tell us the workload, we’ll tell you the box

The same recommendations our engineers give on sales calls, minus the call. The runner-up is listed because honest sizing beats upsells.

What are you running?

Bursts CPU on signals, idles between slots. Network-bound 95% of the time.

Latency to RPC/gRPC dominates PnLCPU mostly idleStateless or near-stateless

Our recommendation

vps.pro8 vCPU · 32 GB · 320 GB NVMe

$190/mo

The bot spends its life waiting on the wire. 8 shared vCPU covers the duty cycle with room to spare, and the money stays in the strategy.

Sanity-check with an engineer

Step up ifmpstat shows steal > 1% in peak hours, or your loop decodes every transaction.

vds.standard · $245/mo

Operational maturity

How the fleet is run

The unglamorous part of infrastructure is the part you are actually buying. Policies below apply fleet-wide and are written into the SLA.

§Capacity policy

·VPS pools run fair-share vCPU scheduling with headroom held in reserve
·VDS and metal capacity is never shared; pinned capacity is held hot
·Spare capacity stays racked and powered, so orders never wait for a host to be carved

§Hardware lifecycle

·AMD EPYC platforms, DDR4 ECC across the fleet
·Datacenter-grade storage with redundancy (mdraid1) on virtualized pools
·Every metal box burn-in tested before handover: CPU, RAM, drive surface, NIC

§Failure & recovery

·VPS: nightly snapshots, restore to a healthy host < 15 min
·VDS: live migration off a failing host, target < 8 min
·Metal: IPMI / KVM-over-IP always on, hardware MTTR < 4 h in the business window

§Network operations

·Committed QoS bandwidth on VDS; 10–100 Gbps dedicated ports on metal
·Dedicated private VLAN to the NLN RPC / gRPC / WS fleet included on every VDS and bare-metal server
·Fair-share scheduler on shared uplinks; one tenant cannot saturate a host

§Access & control

·Full root. No locked kernels, no restricted outbound ports, no inspection
·Key-only SSH from first boot; password auth ships disabled
·Bring your own ISO: Ubuntu, Debian, Rocky, Arch, NixOS all routine

§SLO & credits

·99.99% monthly rolling, fleet-wide
·Credits applied automatically on breach, no ticket required
·Only enforcement on your box: outbound SMTP abuse limits

Frequently asked questions

Mechanically, nothing. It's still KVM virtualization on shared hardware. The difference is what it's sized and supported for: Solana-ready OS images, tiers shaped around bot and indexer duty cycles, a provider that also runs the RPC, gRPC, and WebSocket fleet your code will talk to, and a straight upgrade path to VDS and bare metal when your workload outgrows shared vCPU. One vendor, one bill, one support queue that understands Solana workloads.

Related products

Solana VDS

Step up from VPS when you need guaranteed CPU cores and RAM, no hyperthread sharing.

Bare-metal Solana servers

Full physical machines for validators, MEV pipelines, and 128+ GB RAM workloads.

Solana RPC nodes

The endpoint your VPS will spend most of its time hitting. Same provider, one bill.

Yellowstone gRPC nodes

Streaming layer for sniper bots and MEV searchers running on VPS hardware.

ez-wallet API

Programmatic Solana wallet creation. Pairs well with VPS-hosted custody and CEX onboarding flows.

Geyser Plugin Hosting

Managed Geyser plugin slots on production validators.

From order to running bot, same day

vps.starter $100/mo · vps.pro $190/mo · vps.enterprise $350/mo. Validator-adjacent in Frankfurt on a private VLAN to the NLN RPC, gRPC, and validator fleet. Root SSH same day, most orders within the hour during EU business hours. Full root, your choice of OS.

Talk to an engineer

Custom images, extra IPs, or bespoke regions on committed contracts? An engineer answers, not a quota-carrying rep.

order → root ssh

01
Create an account
Email and a password. No card required to look around.
02
Pick tier + paste your SSH key
The exact specs and prices on this page. No checkout surprises.
03
Root SSH lands in your inbox
Key-only auth, your chosen OS image, ready for the NLN data fleet.

Shared Compute · VPSFrankfurt · validator-adjacent

Solana VPS in Frankfurt, sized for trading workloads

Configure a server Talk to an engineer

vps.pro · first five minutes

$ ssh root@185.x.x.x
Welcome to Ubuntu 24.04 LTS · vps.pro · fra
$ apt install -y nodejs npm && npm i @solana/web3.js
added 42 packages in 6s
$ systemctl enable --now sniper-bot.service
Created symlink → /etc/systemd/system/sniper-bot.service
$ journalctl -u sniper-bot -f
slot 367201442 · processed
slot 367201443 · processed
# order to running bot, same day

Provisioning

Same-day

most within the hour, EU business hours

Access

Full root

KVM · any OS

NLN fleet path

<0.5 ms

private VLAN to RPC/gRPC

Uptime SLO

99.99%

monthly · automatic credits

from $100/mo

Catalog · 3 tiers

Configure a VPS

Every spec and price below is exactly what hits checkout. The catalog is generated from the same table that does the billing.

Need a shape that isn't on the rail: more drives, GPUs, 100 Gbps? We quote custom builds in 24 h.

vps.pro

The production default for trading bots that wait on the network.

$190/mo

monthly · cancel anytime

Compute

8 vCPU · shared EPYC

8 vCPU · fair-share scheduled on shared EPYC

Memory

32 GB DDR4 ECC

dedicated allocation, shared channels

Storage

320 GB NVMe

▮ 320 GB NVMe

local NVMe (host mdraid1) · nightly snapshots · restore < 15 min

Network

1 Gbps fair-share2 TB included egress

Contract

·Own kernel, own IP, full root
·Fair-share vCPU and network scheduling
·Local NVMe storage, nightly snapshots
·Key-only SSH from first boot, any OS
·Provisioning: Same-day · most within the hour (EU business hours)
·SLO: 99.99% monthly · automatic credits

Sized for

Production trading botsLight indexersSniper bots (bursty)

Not for: Hot decode loops at full gRPC volume: that is vds.standard territory.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee

What this is

A working box for chain-adjacent workloads

Tenancy, stated honestly

Shared vCPU, fairly scheduled, with a one-liner to measure steal time yourself and a clear trigger for when to move up.

Measure, don’t trust

We don’t publish latency numbers for VPS. Your first-boot checklist benchmarks the box in five minutes; your numbers are the real ones.

One vendor for the stack

The RPC, gRPC, and WebSocket fleet your bot consumes is ours too. Hosting and data on one bill, with an upgrade path to VDS and metal.

The fine print, in large print

What shared vCPU means, and when to leave it

CPU steal time

real, measure it with mpstat

Cache & memory bus

shared with neighbors

Failure blast radius

host fails → snapshot restore < 15 min

You pay for

a working box, not exclusivity

yours time-sliced other tenants host reserve

measure it yourself · run during peak hours

# Steal-time check: 5 samples, all cores
mpstat -P ALL 1 5 | awk '/Average/ \
  {print "CPU " $2 " steal=" $6}'

# healthy VPS:   steal < 2%
# move to VDS:   steal > 3% sustained

We put the upgrade trigger in the docs on purpose. Mis-sold workloads churn; measured ones stay.

Sizing, consultatively

Tell us the workload, we’ll tell you the box

The same recommendations our engineers give on sales calls, minus the call. The runner-up is listed because honest sizing beats upsells.

What are you running?

Bursts CPU on signals, idles between slots. Network-bound 95% of the time.

Latency to RPC/gRPC dominates PnLCPU mostly idleStateless or near-stateless

Our recommendation

vps.pro8 vCPU · 32 GB · 320 GB NVMe

$190/mo

The bot spends its life waiting on the wire. 8 shared vCPU covers the duty cycle with room to spare, and the money stays in the strategy.

Sanity-check with an engineer

Step up ifmpstat shows steal > 1% in peak hours, or your loop decodes every transaction.

vds.standard · $245/mo

Operational maturity

How the fleet is run

The unglamorous part of infrastructure is the part you are actually buying. Policies below apply fleet-wide and are written into the SLA.

§Capacity policy

·VPS pools run fair-share vCPU scheduling with headroom held in reserve
·VDS and metal capacity is never shared; pinned capacity is held hot
·Spare capacity stays racked and powered, so orders never wait for a host to be carved

§Hardware lifecycle

·AMD EPYC platforms, DDR4 ECC across the fleet
·Datacenter-grade storage with redundancy (mdraid1) on virtualized pools
·Every metal box burn-in tested before handover: CPU, RAM, drive surface, NIC

§Failure & recovery

·VPS: nightly snapshots, restore to a healthy host < 15 min
·VDS: live migration off a failing host, target < 8 min
·Metal: IPMI / KVM-over-IP always on, hardware MTTR < 4 h in the business window

§Network operations

·Committed QoS bandwidth on VDS; 10–100 Gbps dedicated ports on metal
·Dedicated private VLAN to the NLN RPC / gRPC / WS fleet included on every VDS and bare-metal server
·Fair-share scheduler on shared uplinks; one tenant cannot saturate a host

§Access & control

·Full root. No locked kernels, no restricted outbound ports, no inspection
·Key-only SSH from first boot; password auth ships disabled
·Bring your own ISO: Ubuntu, Debian, Rocky, Arch, NixOS all routine

§SLO & credits

·99.99% monthly rolling, fleet-wide
·Credits applied automatically on breach, no ticket required
·Only enforcement on your box: outbound SMTP abuse limits

Frequently asked questions

Related products

Solana VDS

Step up from VPS when you need guaranteed CPU cores and RAM, no hyperthread sharing.

Bare-metal Solana servers

Full physical machines for validators, MEV pipelines, and 128+ GB RAM workloads.

Solana RPC nodes

The endpoint your VPS will spend most of its time hitting. Same provider, one bill.

Yellowstone gRPC nodes

Streaming layer for sniper bots and MEV searchers running on VPS hardware.

ez-wallet API

Programmatic Solana wallet creation. Pairs well with VPS-hosted custody and CEX onboarding flows.

Geyser Plugin Hosting

Managed Geyser plugin slots on production validators.

From order to running bot, same day

Talk to an engineer

Custom images, extra IPs, or bespoke regions on committed contracts? An engineer answers, not a quota-carrying rep.

order → root ssh

01
Create an account
Email and a password. No card required to look around.
02
Pick tier + paste your SSH key
The exact specs and prices on this page. No checkout surprises.
03
Root SSH lands in your inbox
Key-only auth, your chosen OS image, ready for the NLN data fleet.

Solana VPS in Frankfurt, sized for trading workloads

Configure a VPS

A working box for chain-adjacent workloads

What shared vCPU means, and when to leave it

Tell us the workload, we’ll tell you the box

How the fleet is run

Frequently asked questions

Related products

From order to running bot, same day

Ready to get started?

Solana VPS in Frankfurt, sized for trading workloads

Configure a VPS

A working box for chain-adjacent workloads

What shared vCPU means, and when to leave it

Tell us the workload, we’ll tell you the box

How the fleet is run

Frequently asked questions

Related products

From order to running bot, same day

Ready to get started?