Google Cloud Run

Google Cloud (GCP) Run is designed as an improvement of App Engine, incorporating many of the best features of both App Engine standard environment and App Engine flexible environment. Cloud Run is a fully-managed compute environment for deploying and scaling serverless HTTP containers that enables you to run frontend and backend services, batch jobs, and even host LLMs running on vGPUs.

Google Cloud Run functions

In August 2024, Google merged Cloud Functions into Cloud Run, introducing Cloud Run functions. Functions are built with Cloud Build and deployed as Cloud Run services, with access to Cloud Run features.

Focus

Cloud Run allows you to run code either as a continous service that responds to web requests and events, or as a job that exits when finished. Both operate within the same environment and can use the same integrations. You can write code using your favorite language, framework, and libraries, package it up as a container, run gcloud run deploy and your app will be live. Or, you can directly deploy from source and let Cloud Run handle containerizing your code with buildpacks . Cloud Run partly mitigates vendor lock-in with standard OCI container images, a declarative API model and resources that can be represented in YAML files.

Key Features

Cloud Run can access most major Google Cloud services using language-specific API client libraries and/or REST APIs, including Cloud CDN for Edge computing. Microservices deployed in Cloud Run scale automatically based on the number of incoming requests.

An extensive overview of the quotas and limits can be found in the Cloud Run Resources.