FunctionPriceBook

Cold starts and how serverless pricing models differ

By Editorial team · 2026-06-02

In short: Scale-to-zero is why serverless is cheap when idle — but the first request after idle pays a cold-start latency: milliseconds on V8-isolate edge runtimes (Cloudflare Workers, Deno Deploy), 1-2 seconds on Lambda/Cloud Run containers, up to 30-60 seconds on Render's free tier. Billing model matters too: wall-clock GB-second platforms (Lambda, Azure) charge you for I/O wait; CPU-time platforms (Workers, Vercel active CPU) don't. Paying to keep an instance warm trades cost for latency.

Cold starts and pricing are two sides of the same coin. The reason serverless costs $0 while idle is that it scales to zero — and the price of scaling to zero is that the next request has to wake something up. How much that costs you, in both latency and money, depends on the platform.

Cold-start latency by runtime type

Runtime typeExamplesTypical cold start
V8 isolates (edge)Cloudflare Workers, Deno Deploy~milliseconds
Container FaaSAWS Lambda, Cloud Run, Azure Functions~0.2-2 seconds
Free spin-down instancesRender free tier30-60 seconds

Snapshot captured June 2026.

How billing model changes the cost of a cold start

This is the same wall-clock-vs-CPU-time distinction that makes I/O-bound code cheaper on Workers than Lambda (see the true cost comparison).

Paying to remove cold starts

Every major FaaS lets you keep instances warm — provisioned concurrency (Lambda), min instances (Cloud Run), always-ready (Azure Flex Consumption). They work, but they remove scale-to-zero and add a fixed cost, so you’re back to paying for idle time. That’s the fundamental trade-off:

Scale-to-zero = cheapest when idle, slowest first request. Warm instances = fastest, but you pay for idle.

See which platforms scale to zero and which bill while running.

Bottom line

If latency on the first request matters, prefer an edge runtime (millisecond cold starts) or pay to keep a minimum warm. If cost matters more than tail latency, embrace scale-to-zero. Either way, model the billed duration — including init time on wall-clock platforms — in the calculator.

Frequently asked questions

Do you pay for serverless cold starts?

On wall-clock GB-second platforms like AWS Lambda you pay for the initialisation time as part of the billed duration. On CPU-time platforms like Cloudflare Workers, cold-start cost is negligible because V8 isolates start in milliseconds and you're billed on CPU time, not wall-clock. The bigger cost of cold starts is latency, not money.

How do I avoid serverless cold starts?

Keep a minimum number of instances warm (provisioned concurrency on Lambda, min instances on Cloud Run, always-ready on Azure Flex Consumption) — but that removes scale-to-zero and adds a fixed cost. Alternatively use an edge runtime (Cloudflare Workers, Deno Deploy) whose cold start is milliseconds, so warming is unnecessary.

Related articles

Last updated: 2026-06-02