Free cloud cost tools: two calculators, 117 priced instance pages
No signup: two free calculators and priced AWS & GCP instance pages to check idle cloud waste before you provision.
Idlefy flips the default: servers stay off. Your team starts one for exactly as long as they need — from Slack, Telegram, or Web — and it shuts itself down when the timer runs out. No algorithm guessing. Up to 67% off your cloud bill.
+$100 in credits after upgrading to Pro
No credit card required · Free plan available
Not an idle detector. A different default.
Idlefy is a cloud cost platform built on one rule: development servers stay stopped unless someone actually needs them. Instead of hunting for idle machines after the money is spent, it flips the default — a VM is off until an engineer takes a lease on it. You pick a machine in Slack, Telegram, or the web dashboard, choose how long you need it, and it starts in about a minute. When the lease expires, the server shuts itself down.
Need more time? Extend from the same chat. Idlefy warns you before the timer runs out — and if you're still mid-task at expiry, on Pro the Safety System grants one 30-minute grace period before the force-stop. The timer is always set by a person — Idlefy never decides on its own that a server looks idle. Persistent disks survive every stop, and Idlefy only controls servers you explicitly tag — on AWS that boundary is enforced by the IAM policy itself.
Dev boxes, staging, GPU rigs. Untagged machines stay read-only — never started, never stopped. On AWS the IAM policy itself enforces it.
A managed VM's resting state is stopped. Nobody pays for an idle Friday night again.
A tap in Slack, a message in Telegram — or one click on the web — starts the machine for the hours you choose. Extend anytime.
At expiry the VM stops on its own — still working? On Pro the Safety System grants one 30-minute grace first. Persistent disks stay.
A free idle audit shows what 24/7 servers really cost you — before you change anything.
One cloud connection. Idlefy sees all instances but can control only ones YOU tag — on AWS that's verifiable in the IAM policy, not a promise.
A month of usage history, analyzed read-only from data your cloud already stores. Nothing to install — and the audit never touches a running machine.
Every idle hour, priced. Per instance, per hour of the week. Then put the worst offenders on the lease schedule.
The audit finds the waste. Shutdown stays on the lease timer your team sets — never on a guess.
Real results — 17 servers, 141 days of data
Seamless integration with your existing tools
Full Virtual Machine management from your favorite messenger
Your Virtual Machine data-server rental expires in 20 min
App Home dashboard with one-click actions
Active rentals: 3
This month saved: $240
Native desktop widget for instant Virtual Machine access
We don't ask for admin rights. Only the narrow, tag-scoped permissions the product actually needs.
OIDC (Web Identity Federation) on AWS, Workload Identity Federation on GCP — keyless on both, automatic token rotation, no access keys to manage.
On Pro, connect Okta, Auth0, or another compatible OIDC provider and keep joiners and leavers where they already live. Domains are proven by DNS before they grant anything, and SSO can be made mandatory. How it works.
{
"Sid": "IdlefyVMManagement",
"Effect": "Allow",
"Action": [
"ec2:StartInstances",
"ec2:StopInstances",
"ec2:RebootInstances"
],
"Resource": "*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/idlefy": "enabled"
}
}
} Start saving immediately. Upgrade only when you grow.
For Sandbox & Personal Use.
For Growing Dev Teams.
For Scale & Security.
Free calculators and real instance prices — no signup, no email gate.
Hourly and monthly prices for 135 AWS and GCP instance types — and what each one wastes when it runs 24/7.
Browse all instance pricesNotes on cloud cost, FinOps, and dev infra.
No signup: two free calculators and priced AWS & GCP instance pages to check idle cloud waste before you provision.
DevOps shifted from shipping speed to cost control. A case study where one team cut up to 67% of dev VM spend by killing idle infra.
A forgotten Friday deployment burned $5K over a weekend. The fix wasn't a smarter alarm — it was flipping the default. The origin story behind Idlefy.