Every Alchemy app on GCP needs somewhere for its code to run. Each
runtime takes the same Effect program: you bundle it with main, bind
resources to it, and Alchemy grants the IAM.
- Cloud Run service (
GCP.Function) — an HTTP server that scales to zero. Use this by default. - Cloud Run job (
GCP.Run.Job) — a batch task that runs to completion and exits. - Cloud Run worker pool (
GCP.Run.WorkerPool) — always-on containers with no URL, for queue consumers. - Cloud Function (
GCP.CloudFunctions.Function) — an HTTP function built in the cloud, with no local Docker. - GKE (
Kubernetes.DeploymentonGCP.Container.Cluster) — Kubernetes workloads on a cluster you own.
At a glance
| Service | Job | Worker pool | Cloud Function | GKE | |
|---|---|---|---|---|---|
| Entry | fetch | run | run or a pull event source | fetch | fetch or run |
| Triggered by | HTTP, push event sources | RunJob, Cloud Scheduler | Nothing: always running | HTTP, push event sources | HTTP via a load balancer, Jobs on apply |
| Cost shape | Per request; zero idle | Per execution; zero idle | Per running instance | Per request; zero idle | Cluster plus pods |
| Packaging | Image built locally | Image built locally | Image built locally | Source built by Google | Image built locally |
| Deploy time | About a minute | About a minute | About a minute | A few minutes | Cluster: ~10 min. Pods: seconds |
Cloud Run service — the default
An HTTP container that scales with traffic and down to zero. It serves
many requests per instance, streams responses, and runs your fetch
handler as-is. Pub/Sub, Cloud Storage, Cloud Scheduler, and Eventarc
can all push to it.
Cloud Run job — batch work
A container with a run entry. It starts, does the work, and exits.
Nothing is billed between executions. Start it from a service with the
RunJob binding, or on a timer with Cloud Scheduler.
Cloud Run worker pool — background consumers
A pool of containers with no URL. It suits a worker that pulls from Pub/Sub around the clock and should keep running whether or not anyone sends HTTP traffic.
Cloud Functions — no Docker needed
A 2nd-gen Cloud Function runs on Cloud Run underneath. The difference is the build: Alchemy uploads your bundled program and Google builds it, so you can deploy from a machine without Docker.
GKE — Kubernetes
Reach for GKE when you need Kubernetes itself: Helm charts, long-lived pods, custom networking, or workloads you already run elsewhere. Bindings work the same way. They grant roles to the pod's identity through Workload Identity.
→ GKE
Where next
- Cloud Run — the runtime the rest of these docs build on.
- How bindings grant IAM — what happens when a runtime binds a resource.