Cloudflare Workers vs AWS Lambda
Both run your code without a server to manage. Workers run at the edge on a V8-isolate model with near-zero cold starts; Lambda runs functions in AWS regions with a container model and the full AWS ecosystem behind it. The trade is edge-simplicity versus ecosystem-depth.
| Dimension | Cloudflare Workers | AWS Lambda |
|---|---|---|
| Where it runs | Edge (hundreds of locations) | Chosen AWS region(s) |
| Cold starts | Effectively none (isolates) | Noticeable (containers) |
| Runtime model | Web-standard APIs, limits | Full Node/other, more headroom |
| Storage | D1, KV, R2, Durable Objects | RDS, DynamoDB, S3, everything |
| Ecosystem | Focused, fast to start | Vast, deep, more complex |
| Idle cost | Rounds to zero at low traffic | Low, but region + services add up |
Pick Cloudflare Workers when
Latency-sensitive apps, sites and APIs, anything global by default, and small teams who want near-zero ops and idle cost. It is what I run Makmur Motor, Venmail and this site on.
Pick AWS Lambda when
Workloads already deep in AWS, long-running or heavy compute, or when you need a specific AWS service (SQS, Step Functions, big RDS) as the backbone.
Verdict
For edge-native apps, sites and lean APIs, Cloudflare Workers wins on cold starts, latency and idle cost with far less operational surface. For heavy, long-running compute or an app already living inside AWS, Lambda plus the AWS ecosystem is the pragmatic choice. Pick the platform your data gravity already sits in.