Serverless CPU / functions pricing comparison
10 results
| Provider | Plan | Price | Regions | Visit | ||
|---|---|---|---|---|---|---|
|
|
Modal CPU | $5.10 | - |
$0
/1M req
Verified
|
3 countries | Visit → |
|
|
Serverless Functions | $1.67 | 400,000 |
$0.150
/1M req
Verified
|
2 countries | Visit → |
|
|
AWS Lambda | $16.67 | - |
$0.200
/1M req
Verified
|
3 countries | Visit → |
|
|
Azure Functions | $16.00 | - |
$0.200
/1M req
Verified
|
3 countries | Visit → |
|
|
Cloudflare Workers | $12.50 | 3,000,000 |
$0.300
/1M req
Verified
|
3 countries | Visit → |
|
|
Cloud Functions (2nd gen) | $2.50 | - |
$0.400
/1M req
Verified
|
3 countries | Visit → |
Fastly
|
Fastly Compute | $3.50 | - |
$0.500
/1M req
Verified
|
3 countries | Visit → |
|
|
Vercel Serverless Functions | $18.00 | 1,000,000 |
$0.600
/1M req
Verified
|
3 countries | Visit → |
|
|
Deno Deploy Pro | $12.50 | 1,000,000 |
$2.00
/1M req
Verified
|
3 countries | Visit → |
|
|
Netlify Functions | $8.60 | 125,000 |
$2.50
/1M req
Verified
|
3 countries | Visit → |
Frequently asked questions
How is serverless CPU priced?
You pay per request and per GB-second of memory used while your code runs, so idle time is free. The table shows both the per-request and per-GB-second rates to compare providers.
What is the cheapest serverless platform?
It depends on your request volume and execution time. Sort by per-request and per-GB-second cost; high-volume short functions favor low request fees, while long-running ones favor low compute rates.
Do serverless platforms have a free tier?
Many include a generous monthly free allowance of requests and compute, enough to run small workloads at no cost. Where a provider publishes one it is shown; it is a great way to start before paying.
What are cold starts and do they cost extra?
A cold start is the delay when a new instance spins up after idle. It is not billed directly but adds latency; pick a provider with fast cold starts for user-facing endpoints.
Is serverless cheaper than a VPS?
Serverless wins for spiky or low-volume workloads because you pay nothing when idle; a VPS is cheaper for steady, always-on load. Compare against VPS pricing on our VPS page for your traffic pattern.
What is the maximum function duration?
Platforms cap how long a single invocation can run, from seconds to many minutes. Check the limit for long jobs; batch or background work may need a container or VPS instead.