Cloudflare Workers is a serverless execution environment that allows you to create entirely new applications or augment existing ones without configuring or maintaining infrastructure.

1 detections
1 websites tracked
Updated 26 May 2026

Websites Using Cloudflare Workers

What Are Cloudflare Workers?

Cloudflare Workers is a serverless compute platform that runs your code directly on Cloudflare's global edge network, in data centers close to your users, rather than in a single central region. Instead of provisioning servers or containers, you write a small JavaScript (or WebAssembly) function, deploy it, and Cloudflare executes it at the edge in response to incoming requests. Because the code runs at the same network locations that already serve cached content and proxy traffic, Workers can intercept, transform, and respond to requests with very low latency from almost anywhere in the world.

Cloudflare introduced Workers in 2017 as part of its broader move from being purely a CDN and security provider into an edge computing platform. It is widely regarded as one of the pioneering and most prominent edge-serverless offerings, and it anchors a growing suite of edge products. The platform is built on a distinctive architecture: rather than spinning up a container or virtual machine per function, Workers run inside lightweight V8 isolates, the same sandboxing technology that separates tabs in the Chrome browser, which allows enormous numbers of small functions to run efficiently with minimal startup overhead.

It is worth being precise about what Workers is. It is not a traditional web host where your whole site's files live, and it is not a browser tool or plugin. It is a place to run code, your application logic, at the edge of Cloudflare's network. A Worker might serve an entire site, act as an API, modify requests and responses flowing through Cloudflare, perform authentication, run A/B tests, or stitch together data from multiple back ends. Increasingly, full applications and frameworks deploy to Workers, and Cloudflare pairs the compute layer with complementary services for storage, databases, and static assets.

From a detection standpoint, the important nuance is that Workers run within Cloudflare's edge, which already fronts a vast number of sites. So evidence that a site is on Cloudflare is common, while evidence that it specifically runs Workers is more subtle and depends on how the Worker is deployed and what it does. Recognizing the difference between "this site uses Cloudflare" and "this site runs code on Cloudflare Workers" is central to interpreting the signals correctly.

How Cloudflare Workers Work

The defining technical choice behind Workers is the use of V8 isolates instead of containers. An isolate is a lightweight, sandboxed execution context within a single process; many isolates can share one runtime while remaining isolated from one another. This design means a Worker can start almost instantly, avoiding the cold-start delay associated with spinning up a container or virtual machine, and it allows Cloudflare to run a given Worker in every data center in its network without dedicating heavy resources to each one. The trade-off is that Workers operate within a web-standard runtime built around the same APIs browsers expose, rather than a full server operating system.

A Worker is fundamentally a request handler. It receives an incoming HTTP request as a standard Request object and returns a Response. Within that handler, the code can do a wide range of things: return a generated response directly, fetch from an origin server or another API and transform the result, rewrite headers, route based on the URL, set cookies, run authentication or authorization logic, cache intelligently, or assemble a page from multiple sources. Because this all happens at the edge, the work occurs close to the user and before traffic ever needs to reach a distant origin.

Workers rarely operate alone. Cloudflare provides a family of edge services that Workers bind to: a key-value store for low-latency reads, an edge SQL database, object storage with no egress fees, durable coordination primitives for stateful logic, queues, and a static-asset hosting layer for deploying full sites and front-end applications. A modern application built on this stack might serve its static front end from the asset layer, run its API and server-side logic in Workers, and read and write data through the edge database and storage, all within Cloudflare's network.

The development and deployment workflow centers on Cloudflare's command-line tooling, which lets developers build, test locally, and publish Workers, and on configuration that maps routes or domains to specific Workers. When a request arrives at Cloudflare for a route handled by a Worker, the network executes the Worker in the nearest data center, and the Worker's response is returned to the user, often without the request ever touching the customer's origin at all.

This architecture, edge execution, isolate-based concurrency, and tight integration with edge storage, is what makes Workers fast and scalable. It also shapes how the platform appears from the outside: a Worker-served response comes from Cloudflare's edge and carries Cloudflare's characteristic markers, while the application logic itself is invisible, executed server-side at the edge and reflected only in the response it produces.

How to Tell if a Website Uses Cloudflare Workers

Detecting Cloudflare Workers specifically is more nuanced than detecting most technologies, because Workers run inside Cloudflare's edge, which already fronts an enormous share of the web. StackOptic inspects response headers and behavior from the server side, and the signals below distinguish "on Cloudflare" from "running Workers," with appropriate honesty about the limits.

Cloudflare's baseline presence. First, confirm Cloudflare itself. The clearest tells are a Server: cloudflare response header and a cf-ray header on responses, both of which Cloudflare adds to traffic it proxies. These indicate the site is behind Cloudflare, which is a prerequisite for Workers but does not by itself prove Workers are in use.

Workers.dev subdomains. Workers can be deployed to a free *.workers.dev subdomain. If a URL or a referenced endpoint uses workers.dev, that is direct evidence of a Cloudflare Worker, since that domain exists specifically to host Workers.

Worker-served responses without a separate origin. When a Worker generates the entire response, the content is served from Cloudflare's edge with Cloudflare headers and no indication of a distinct origin server behind it. A dynamic, application-like response that nonetheless carries only Cloudflare's Server and cf-ray headers, with no separate origin server header, is consistent with a Worker handling the request, though it is not absolute proof.

Cloudflare Pages and asset patterns. Sites deployed through Cloudflare's static-asset and Pages products, frequently paired with Workers for dynamic logic, can expose related headers and hostnames. These reinforce that the application is running on Cloudflare's edge platform rather than merely being proxied through its CDN.

Here is how to check each signal yourself:

MethodWhat to doWhat it reveals
curl -Icurl -I https://example.comServer: cloudflare and a cf-ray header confirm Cloudflare; basis for Workers
Browser DevToolsNetwork tab, inspect response headerscf-ray, Cloudflare headers, and whether any separate origin header appears
URL inspectionCheck for *.workers.dev or Pages hostnamesDirect evidence of a deployed Worker or Cloudflare Pages site
WappalyzerRun the extension on the live pageIdentifies Cloudflare, and sometimes Workers/Pages, when exposed
BuiltWithLook up the domainCloudflare usage and related edge-platform signals

A quick check is curl -sI https://example.com | grep -iE "server|cf-ray". Seeing Server: cloudflare and a cf-ray value confirms Cloudflare; a workers.dev hostname confirms a Worker specifically. For methodology, see our guides on how to tell if a website uses Cloudflare or another CDN and how to read a website's HTTP headers.

The honest caveat here is the most important part. Cloudflare's edge masks the origin and stamps the same headers on ordinary proxied sites and on Worker-powered ones alike, so Server: cloudflare and cf-ray confirm Cloudflare but cannot, on their own, prove that Workers are running. A site might be a static page behind Cloudflare's CDN, or it might be a full application executing in Workers, and both can look similar from a single header check. The strongest positive signals are a workers.dev hostname or a clearly dynamic response served entirely from Cloudflare's edge with no separate origin. Absent those, the most accurate statement is often that a site is on Cloudflare and may be using Workers, rather than a definitive yes. This is exactly the kind of distinction where combining signals, headers, hostnames, response behavior, and where server-side analysis of the raw response helps, because it reads what Cloudflare actually returns without browser interference. To understand how Workers relate to the wider platform, see Cloudflare.

Key Features

  • Edge execution on V8 isolates. Code runs in lightweight isolates across Cloudflare's global network with minimal cold-start overhead.
  • Request and response handling. Workers intercept, transform, route, and generate HTTP responses at the edge before traffic reaches an origin.
  • Integrated edge storage. Bindings to a key-value store, an edge SQL database, object storage, durable stateful primitives, and queues.
  • Static asset and full-stack hosting. Deploy entire front ends and applications to the edge alongside Worker logic.
  • WebAssembly support. Run compiled languages via WebAssembly in addition to JavaScript and TypeScript.
  • Global low-latency reach. Code runs close to users in data centers worldwide rather than in a single region.
  • Developer tooling. A command-line workflow for local development, testing, and deployment, with route and domain mapping.

Pros and Cons

Pros

  • Extremely low latency by running logic at the edge near users.
  • Near-instant startup thanks to isolate-based execution, avoiding container cold starts.
  • Scales automatically across a global network with no servers to manage.
  • Tight integration with edge storage and asset hosting enables full applications on one platform.

Cons

  • A web-standard runtime with constraints that differ from a full server environment, so not every Node.js library or workload fits directly.
  • Execution limits and an edge-oriented model require designing for short, stateless-leaning request handling.
  • Deep reliance on the platform can increase coupling to Cloudflare's ecosystem.
  • From the outside, distinguishing Workers from ordinary Cloudflare proxying is not always definitive.

Cloudflare Workers vs Alternatives

Workers competes with other serverless and edge compute platforms. The table below clarifies where it fits.

PlatformExecution modelRuns at edgeBest for
Cloudflare WorkersV8 isolatesYes, global edgeLow-latency edge logic, APIs, full-stack edge apps
AWS LambdaContainers/microVMsRegional (Lambda@Edge for edge)Broad serverless workloads in the AWS ecosystem
Vercel Edge / FunctionsEdge runtime and serverlessYes, edge runtimeFront-end frameworks and Jamstack deployments
Deno DeployV8 isolatesYes, global edgeTypeScript-first edge functions
Fastly ComputeWebAssemblyYes, edgeHigh-performance edge compute via Wasm

If you are comparing where to run code at the edge, contrast Workers with a front-end-focused platform like Vercel, or read our overview of what a CDN is and whether you need one to see how edge compute builds on edge delivery.

Use Cases

Cloudflare Workers is most compelling wherever low latency, global reach, and request-level logic matter. Developers use it to build APIs that respond from the nearest edge, to add authentication and authorization in front of an origin, and to rewrite or personalize requests and responses on the fly. Because Workers sit in the request path, they are ideal for tasks like A/B testing, feature flagging, redirects, header manipulation, and bot or abuse mitigation logic.

It also serves as the backbone for full edge applications: a front end deployed to Cloudflare's asset layer, an API and server-side rendering handled by Workers, and data stored in the edge key-value store or SQL database. Teams use it to offload work from origin servers, to stitch together multiple back-end services into a single edge-served response, and to deliver dynamic, personalized content quickly worldwide. For technology research, detecting Workers, or at least Cloudflare's edge platform, signals a modern, performance-minded engineering team.

A few concrete patterns illustrate the range. A SaaS company might run its public API on Workers so customers everywhere get fast responses, with the Worker authenticating requests and caching results at the edge. A media site might use a Worker to personalize content and run experiments without touching its origin on every request. A startup might build its entire product on the Cloudflare stack, front end on the asset layer, logic in Workers, data in the edge database, avoiding traditional servers altogether. The common thread is moving computation to the edge to cut latency and simplify operations.

From a competitive-intelligence perspective, identifying that a site runs on Cloudflare's edge platform, and especially that it uses Workers, indicates a forward-looking technical organization comfortable with serverless and edge architectures. For vendors selling developer tools, infrastructure, or complementary services, that is a useful qualifying signal. The caveat from the detection section bears repeating, though: confirming Cloudflare is easy, while confirming Workers specifically requires stronger evidence such as a workers.dev hostname, so the most honest read is sometimes "on Cloudflare, likely using Workers" rather than a flat certainty.

Frequently Asked Questions

What is the difference between Cloudflare and Cloudflare Workers?

Cloudflare is the overall platform, providing CDN, DNS, and security that proxy and protect a site. Cloudflare Workers is a specific product within it: a serverless compute service that runs your code at Cloudflare's edge. A site can be on Cloudflare without using Workers (just using the CDN and security), but it cannot use Workers without being on Cloudflare. That is why detecting Cloudflare is common and easy, while detecting Workers specifically requires additional evidence.

How can I tell if a site is actually running Workers and not just using Cloudflare?

Look for stronger, Worker-specific signals. A *.workers.dev hostname is direct evidence of a Worker. A clearly dynamic, application-like response served entirely from Cloudflare's edge, with only Cloudflare headers and no separate origin server header, is consistent with a Worker. By contrast, Server: cloudflare and a cf-ray header alone only prove the site is on Cloudflare. When the strong signals are absent, the accurate conclusion is usually "on Cloudflare, possibly using Workers."

What are V8 isolates and why do they matter?

V8 isolates are lightweight, sandboxed execution contexts within a single process, the same technology that isolates tabs in the Chrome browser. Workers run in isolates rather than containers or virtual machines, which lets them start almost instantly and run in every Cloudflare data center without heavy per-function overhead. This avoids the cold-start delays common to other serverless platforms and is the key reason Workers can execute globally at very low latency.

Can you build a whole website on Cloudflare Workers?

Yes. Modern applications deploy their static front end to Cloudflare's asset and Pages layer and run their dynamic logic and APIs in Workers, with data in the edge key-value store, SQL database, or object storage. Many web frameworks now support deploying to this stack. The result is a full application that runs at the edge with no traditional servers to manage, though the platform's web-standard runtime and execution limits mean some server-style workloads need adapting.

Does using Workers hide where a site's origin is?

Often, yes. When a Worker generates the full response at the edge, requests may never reach a separate origin, and even when they do, Cloudflare's proxy masks the origin's address behind its own network. The headers you see come from Cloudflare's edge, not the underlying server. This is consistent with how Cloudflare works generally, the edge stands in front of the origin, which is why detecting the origin from the outside is difficult while detecting Cloudflare itself is easy.

Want to identify Cloudflare Workers and the rest of a site's edge and hosting stack? Run any URL through StackOptic at https://stackoptic.com.