Frequently asked questions
What is Azure Container Apps?
Azure Container Apps is a fully managed, serverless container platform for running microservices, APIs, event-driven jobs, and AI apps and agents. It runs your containers on a Kubernetes-based foundation with built-in support for KEDA autoscaling, Dapr, and Envoy — without requiring you to provision, patch, or operate a cluster. Bring a container image from any registry; the platform handles scaling (including scale to zero), ingress, certificates, and revisions.
What are some common uses of Azure Container Apps?
Common workloads include public or internal HTTP and gRPC APIs, microservices that communicate over service discovery or Dapr, event-driven processing and background workers that scale on queue or event triggers, on-demand and scheduled jobs, and AI apps and agents — including MCP servers and model inference on serverless GPUs. See use cases to learn more.
Does Azure Container Apps support scale-to-zero pricing?
Does Azure Container Apps have a free tier?
Yes. On the Consumption plan, your first 180,000 vCPU-seconds, 360,000 GiB-seconds, and 2 million requests each month are free. Usage beyond the free grant is billed per second. See pricing.
What's the difference between active usage and idle usage?
On the Consumption plan, a replica is billed at the active rate while it's starting up or processing requests and using its allocated CPU. When a replica is running but not actively using CPU — for example, kept alive by a minimum-replica setting between requests — it's billed at the lower idle rate. See billing.
What is Azure Container Apps Express?
Azure Container Apps Express is the fastest path from a container image to a running app. It streamlines environment setup so you can go from image to public endpoint in seconds, with sensible defaults instead of upfront configuration. See Get started with Container Apps Express.
Why use Azure Container Apps Sandbox?
Container Apps Sandboxes provide secure, isolated, ephemeral microVM compute with sub-second startup — ideal for running untrusted or AI-generated code at runtime. They contain risk by keeping each execution isolated from your data and shared infrastructure. They also scale to zero when idle. See the Sandboxes docs.
How can I use Azure Functions in an Azure Container Apps environment?
You can host Azure Functions directly inside a Container Apps environment, combining the Functions programming model and triggers with Container Apps networking, scaling, and Dapr integration. This lets event-driven Functions run alongside your other container apps in the same environment. See Azure Functions on Container Apps.
Why use serverless GPUs?
Serverless GPUs provide on-demand GPU compute for AI model inference and image generation without requiring you to provision or manage GPU infrastructure. You scale to zero when idle, so you only pay while workloads run, and your data stays within your environment for built-in governance. See serverless GPUs.