Serverless
Running code without managing servers the platform spins functions up on demand and scales them automatically.
In serverless architectures, you deploy functions (or containers) and the platform decides when to run them, on which machine, and at what scale. AWS Lambda, Vercel Functions, Cloudflare Workers, Google Cloud Run, and Azure Functions are typical examples.
The upsides: pay per execution, scale to zero, no infrastructure to babysit. The trade-offs: cold starts on the first invocation after idle, runtime constraints, and a different mental model than long-running servers.
In 2026 serverless is a default for most new web work. Container-based serverless (Cloud Run, Fly Machines) has narrowed the gap with traditional servers, while edge runtimes have pushed serverless even closer to users.