Fly.io pricing in 2026: what the $5 plan everyone quotes actually costs now
Updated on August 20, 2026
On this page
_Note, August 20, 2026: the fuller and continuously maintained version of this teardown lives at Fly.io Pricing in 2026: The Real 30-Day Bill, which carries three worked bills and the reservation math. Compute rates on this page were re-verified against Fly.io's resource pricing docs{target=blank} on this date.
Quick Answer (2026)
Fly.io in 2026 is pure pay-as-you-go: you pay per running second of a Machine, plus storage, bandwidth, and an optional support plan. A shared-cpu-1x 256MB machine running 24/7 is about $2.02/month of compute. A small production performance-1x 2GB machine is about $32.19/month of compute. There is no free tier for new customers anymore, and the widely quoted "$29/month plan" is a support tier, not a hosting plan. Reserve steady machines for 40% off; leave spiky ones on-demand.
What actually changed
Fly.io retired named subscription plans and moved fully to usage billing. The old free allowance (up to 3 shared-cpu-1x 256MB machines and 3GB of volume storage) is now marked legacy and applies only to grandfathered accounts. A new account pays from the first second a Machine is running. That single change is why so much of the internet's Fly.io advice is now wrong.
Here is the gap between what search results still say and what the meter charges in 2026.
Scroll to see more
| What you still read | What 2026 actually is |
|---|---|
| "$5/mo Hobby plan" | Retired 2024 legacy plan; no subscription compute tier exists |
| "$29/mo Launch plan" | $29 is now the Standard support tier, not hosting |
| "Free tier for everyone" | Free allowance is legacy-only; new accounts pay from second one |
| "Flat monthly server price" | Per-second Machine billing plus storage, bandwidth, support |
The real 2026 meter, decoded
Compute is billed per second while a Machine is started. Converting the per-second rates to a 24/7 month (about 2,592,000 seconds) makes them legible:
Scroll to see more
| Machine preset | Per-second | ~24/7 per 30 days |
|---|---|---|
| shared-cpu-1x, 256MB | $0.00000078 | ~$2.02 |
| performance-1x, 2GB | $0.00001242 | ~$32.19 |
| performance-16x, 32GB | $0.00030912 | ~$801 |
Extra RAM is a separate line: about $5 per 30 days per additional GB on top of the preset. Then the rest of the meter:
- Volumes: $0.15/GB per month of provisioned capacity (you pay for what you allocated, not what you filled).
- Volume snapshots: $0.08/GB per month, first 10GB free each month.
- Stopped-machine storage: about $0.15 per GB of rootfs for a machine stopped 30 days.
- Bandwidth (egress): $0.02/GB in North America and Europe, $0.04/GB in Asia Pacific, Oceania and South America, $0.12/GB in Africa and India.
- Dedicated IPv4: $2/month.
- Support (optional): Standard $29/mo, Premium $199/mo, Enterprise from $2,500/mo. HIPAA workloads add $99/mo.
All figures above are from Fly.io's own resource pricing docs and pricing page, read August 2026.
Trap 1: scale-to-zero is not scale-to-free
Fly Machines can auto-stop when idle, and teams treat that as a $0 state. It is not. A stopped machine bills nothing for compute, but any attached volume keeps billing at $0.15/GB per month of provisioned size, and the stopped rootfs still costs about $0.15/GB for 30 days.
A "scaled to zero" service with a 3GB volume is not $0. It is about $0.45 to $0.60/month of storage doing nothing. Multiply that across a dozen sleeping preview environments and the idle bill is real.
Trap 2: the $29 is support, not the server
On a small Fly.io bill, the largest single line is often the support plan, not the compute. Standard support is $29/month, which is more than fifteen shared-cpu-1x machines running 24/7. Support is genuinely useful once you are in production, but pricing it as if it were the hosting cost is how people conclude "Fly.io is expensive" when the compute underneath is a couple of dollars.
The lever: reserve steady machines for 40% off
Fly.io gives a 40% discount when you reserve a block of compute time instead of paying on-demand. Their own examples: $36/year buys $5/month of usage (shared), and $144/year buys $20/month of usage (performance). The crossover is simple.
- Machine runs 24/7 with steady load -> reserve it. 40% off compute is free money on a workload you were going to run anyway.
- Machine is spiky or auto-stops -> stay on-demand. Reserving a machine you keep stopping wastes the reservation.
Trap 3: egress is priced by geography
The same byte costs six times more leaving Mumbai than Virginia. Serving 100GB/month of egress from North America is $2.00. Serving the identical 100GB from Africa or India is $12.00. For a globally distributed app, where your traffic lands is a pricing decision, not just a latency one.
Three worked 30-day bills
A. Hobby app, one small machine, always on
Scroll to see more
| Line | Cost |
|---|---|
| shared-cpu-1x 256MB, 24/7 | $2.02 |
| 3GB volume | $0.45 |
| 5GB North America egress | $0.10 |
| Total | ~$2.49/mo |
The "$5 Hobby plan" the internet quotes does not exist. You pay about $2.49 of metered usage instead, with no free tier to fall back to.
B. Small production API, humans on call
Scroll to see more
| Line | Cost |
|---|---|
| performance-1x 2GB, 24/7 | $32.19 |
| 10GB volume | $1.50 |
| 50GB North America egress | $1.00 |
| Dedicated IPv4 | $2.00 |
| Standard support | $29.00 |
| Total | ~$64.50/mo |
Support is 45% of this bill. Reserve the machine (40% off compute -> $18.60) and the same setup drops to about $52.10/mo.
C. The "we scaled to zero" preview environment
Scroll to see more
| Line | Cost |
|---|---|
| Stopped machine compute | $0.00 |
| 3GB provisioned volume | $0.45 |
| Stopped rootfs storage | ~$0.15 |
| Total | ~$0.60/mo |
Not $0. Delete the volume, not just stop the machine, if a preview is truly dead.
Where Fly.io fits, and where it does not
Fly.io's per-second model rewards workloads that genuinely start and stop, and punishes always-on machines you forgot to reserve. If your app is a steady 24/7 service, price it against usage-billed 
Sources
- Fly.io, "Fly.io Resource Pricing," https://fly.io/docs/about/pricing/ (read August 2026)
- Fly.io, "Pricing," https://fly.io/pricing/ (read August 2026)
Math check: a 24/7 shared-cpu-1x is about $1.94 of compute, so on most small Fly.io bills the $29 support plan, not the server, is the biggest line. Reserve steady machines for 40% off; leave the spiky ones on-demand.
Written by
Diego AguirreDiego pulls apart cloud and infrastructure invoices line by line so you can see where the money actually goes before the bill lands.
Frequently asked questions
Does Fly.io still have a free tier in 2026?
No. The free allowance (up to 3 shared-cpu-1x 256MB machines and 3GB of volume storage) is now legacy-only and is not offered to new customers. New accounts pay metered usage from the first running second, so budget for real dollars even on a hobby app.
How much does Fly.io actually cost per month in 2026?
A single shared-cpu-1x 256MB machine running 24/7 is about $1.94 of compute per 30 days at the per-second rate. A small production performance-1x 2GB machine is about $31 of compute. The bigger surprise on small bills is usually the optional support plan, which starts at $29/month, not the server itself.
Do you get charged when a Fly.io machine is stopped?
Partly. A stopped machine bills $0 for compute, but you still pay for any attached volume at $0.15/GB per month of provisioned capacity, plus about $0.15 per GB of stopped rootfs for 30 days. Scale-to-zero lowers a bill, it does not zero it.
Is the $29/month a Fly.io hosting plan?
No. In 2026 there are no subscription compute plans. The $29 (Standard), $199 (Premium) and $2,500+ (Enterprise) figures are support tiers layered on top of pay-as-you-go usage. The old $5 Hobby / $29 Launch / $199 Scale plan names still quoted across search results are the retired 2024 legacy pricing.
Related reading
Fly.io Pricing in 2026: The Real 30-Day Bill
Fly.io pricing in 2026 is pure pay-as-you-go: you pay per second for the Machines (micro-VMs) you actually run, not a fixed plan. A small always-on app on a shared-cpu-1x 256MB Machine costs about $2.02 a month; a two-Machine production SaaS with Postgres and Redis lands near $29 a month; a three-region app on performance Machines runs about $226 a month. The old Hobby, Launch, and Scale plan names were retired in the October 2024 pay-as-you-go switch, and the biggest lever left is a machine reservation, which cuts committed compute by 40%.
Railway Pricing in 2026: What You Actually Pay
Railway pricing in 2026 is usage-based: you pay $20 per vCPU and $10 per GB of RAM per month, billed by the second, plus $0.05 per GB of egress and $0.15 per GB of volume storage. Hobby is $5 a month with $5 of usage included; Pro is $20 a month per workspace, not per seat, with unlimited seats and $20 of usage included. There is no permanent free tier, only a 30-day $5 trial, so a small always-on app runs about $5 to $22 a month and a modest production stack near $70.
Render Pricing in 2026: What a Real Monthly Bill Actually Looks Like
Render's advertised prices hide the real bill. In 2026 your cost is three stacked layers: a workspace plan, per-service compute, and metered usage. A solo production app starts near $13/mo (not $7), a small team app around $115/mo. Here is the arithmetic, with three worked 30-day bills.