---
title: "Serverless computing"
description: "Serverless computing is a cloud model where the provider manages servers and you pay only for actual code execution. Learn how it works and where it fits."
url: "https://nodea.io/en/glossary/serverless-computing"
lang: "en"
updated: "2026-07-13"
---

# Serverless computing

> Serverless computing is a cloud model in which the provider fully manages the infrastructure, the developer supplies only code that runs in response to events, and billing is based solely on actual execution time.

**Serverless computing** is a model for executing code in the [cloud](https://nodea.io/en/glossary/cloud) in which the developer supplies only the application logic, while the provider manages all the infrastructure — servers, operating system, scaling and availability. Despite the name, servers still exist; what changes is the programmer's perspective: they stop thinking about machines and think only about code and the events that trigger it.

## How serverless computing works

The most popular realization of this model is **Functions as a Service** (FaaS) — individual pieces of code triggered by an event: an HTTP request through an [API](https://nodea.io/en/glossary/api), a file upload, a database write or a message on a queue. The function starts on demand, performs its task and terminates, while the provider automatically scales the number of concurrent instances with the load — from zero to thousands of invocations and back.

The foundation is **pay-per-use** billing: you pay for actual execution time and the number of invocations, not for a permanently rented machine. When the code is idle, the cost is zero. The price for that convenience is the *cold-start* latency (spinning up an idle function), execution time limits, and tighter coupling to a specific provider's ecosystem.

## Practical application

Serverless is a great fit for event-driven tasks and irregular load: mobile app backends, image processing after upload, form handling, webhooks, scheduled jobs and lightweight APIs. Well-known services include AWS Lambda, Google Cloud Functions and Azure Functions. The model speeds up development because it frees the [DevOps](https://nodea.io/en/glossary/devops) team from worrying about servers, and it handles spiky traffic beautifully. It does not replace everything, though — services with steady, heavy load or a need for full control over the environment often remain faster and cheaper on a dedicated VPS. In practice serverless tends to complement, rather than succeed, traditional infrastructure.

### Are there really no servers in serverless?

Servers still exist — the name refers to the developer's perspective, who no longer has to provision, configure or scale them. The provider manages all the infrastructure behind the scenes. The programmer focuses solely on the code and the application logic, not on the machines it runs on.

### When is serverless a poor fit?

The model can be uneconomical for steady, high and predictable traffic, where a traditional server works out cheaper. Long-running tasks are also problematic due to execution time limits, as is cold-start latency on the first invocation. Applications needing full control over the environment usually run better on a VPS.

## Related

- [Cloud](https://nodea.io/en/glossary/cloud)
- [PaaS](https://nodea.io/en/glossary/paas)
- [IaaS (Infrastructure as a Service)](https://nodea.io/en/glossary/iaas)
- [API](https://nodea.io/en/glossary/api)
- [DevOps](https://nodea.io/en/glossary/devops)

---

Languages: [pl](https://nodea.pl/pl/slownik-pojec/serverless-computing.md)  
Site index: [llms.txt](https://nodea.io/en/llms.txt)