Pricing Documentation
Login Get Started
Build or outsource

Open source on T Cloud: in-house or managed

The patching, the backups, the certificate rotation and the 3am pager do not go away when you self-host. They get assigned to someone. This page puts the two options side by side with a pricing formula you can compute yourself.

Same account, same jurisdiction Formula pricing, no quote needed Scales down as well as up

The work does not disappear. It just moves.

Nobody decides between having operations and not having them. You decide who is accountable, and what that accountability costs in salary, attention and risk.

What running open source in-house actually contains

  • Continuous CVE monitoring across every component you self-host, including the ones nobody remembers installing.
  • OS patching on a schedule that survives a busy quarter, not one that slips the moment a release is due.
  • Version upgrades researched, staged and executed per service, each with its own breaking changes.
  • Backups that are actually restore-tested, which is a different task from backups that run.
  • Certificate rotation before expiry rather than during the outage it causes.
  • 24/7 monitoring and an on-call rota, which requires enough people to make a rota humane.

None of this is difficult work. All of it is relentless, and it competes directly with the roadmap every single week.

In-house operations versus Elestio on your own T Cloud account

Same infrastructure, same account, same jurisdiction. The difference is who carries the operational load.

In-house
Elestio on your T Cloud account
Where the virtual machines run
Your T Cloud account
Your T Cloud account, unchanged
Who holds the Telekom contract
You
You, unchanged
Time to deploy a new service
Days of setup and hardening
One click from 400+ services
Patching and CVE response
Manual, competes with the roadmap
Automated and continuous
Backup verification
Rarely done honestly
Automated with point-in-time recovery
Out of hours incidents
Your on-call rota
24/7 monitoring and intervention
Uptime commitment
Best effort
Up to 99.9% with a 4 hour response
Cost shape
Salaried headcount, fixed
Per resource, hourly, scales down too
Key person risk
High, one engineer knows the stack
Removed, the runbook is the product
If you stop
Nothing changes
Infrastructure stays, you keep root access

What twenty services cost to have managed

The Elestio fee is a formula, so you can compute your own number rather than ask for a quote.

$5
per vCPU
$2.50
per GB RAM
$0.025
per GB disk
5 services
4 vCPU · 8 GB · 160 GB each
$220/mo
$44 per service in management fees
20 services
4 vCPU · 8 GB · 160 GB each
$880/mo
$44 per service in management fees
50 services
4 vCPU · 8 GB · 160 GB each
$2,200/mo
$44 per service in management fees
Management fees only, billed hourly. T Cloud infrastructure is billed separately and directly by Deutsche Telekom on your own account. Support Level 2 adds $50 per service per month for a 99.5% SLA, Level 3 adds $200 for 99.9% and a 4 hour response. Compare the total against the fully loaded cost of the engineering time it replaces at your own rates.

The two arguments that usually decide it

Budget holders start with the monthly number. They sign for different reasons.

Key person risk is the quiet one

In most organizations, one engineer knows how the stack is really wired. When that person takes a new job, the knowledge leaves with them. Handing operations to a provider converts a person-shaped dependency into a contract with a response time attached.

Opportunity cost is the expensive one

Every hour spent patching a wiki or rotating a certificate is an hour not spent on the thing your organization is actually judged on. That cost never appears on any invoice, which is exactly why it goes unmanaged for years.

Reviews

Trusted by 10,000+ Developers Worldwide

Real reviews from real users on Trustpilot.

Frequently Asked Questions

  • Is this cheaper than hiring?

    That depends entirely on your own loaded cost per engineer, which is why the pricing is a formula rather than a quote. Compute your total with $5 per vCPU plus $2.50 per GB RAM plus $0.025 per GB disk across the services you would hand over, add support level if you need an SLA, and compare it against the fully loaded cost of the engineering time it replaces at your rates.

  • Do we lose control by outsourcing operations?

    No. You keep full root access to every server, the virtual machines stay in your own T Cloud account, and your contract with Deutsche Telekom is unchanged. Elestio access runs through a dedicated IAM user that you can revoke from your own console at any time.

  • What if we only want some services managed?

    That works. The fee is charged per managed service based on its resources, so you can hand over the twenty things that are pure operational overhead and keep the ones your team genuinely wants to own.

  • Does the cost scale down if we shrink?

    Yes. Management fees are billed hourly on actual resources, so reducing instance sizes or removing services reduces the fee immediately. Salaried headcount does not behave that way.

  • What happens to our existing Telekom discount?

    It is untouched. Deutsche Telekom bills your infrastructure directly on your existing account, so committed volume and negotiated discounts still apply. Elestio never sits in that billing relationship.

  • What if we want to leave later?

    The infrastructure is already yours and stays where it is. Revoke the IAM user in your own console and the servers keep running unmanaged. You can also clone or migrate managed services to any of the 9 cloud providers Elestio supports.

Compute your number, then decide

Same infrastructure, same account. The only variable is who carries the operational load.

Get Started