> ## Documentation Index
> Fetch the complete documentation index at: https://clubhub.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Google Cloud Run vs Fly

> Why we chose Google Cloud Run instead of Fly

[Fly](https://fly.io) doesn't support auto-scaling down to 0 instances and is far more expensive.
A server running 24/7 with an [open database connection](https://neon.tech/docs/introduction/compute-lifecycle) to [Neon](/stack/neon) will use the corresponding [compute time](https://community.neon.tech/t/compute-time-usage-calculation/649/2).

[Google Cloud Run](/stack/google-cloud-run) supports scaling down to 0 instances and scaling up when requests come in, resulting in no unnecessary open database connections and saved **Neon** [compute time](https://community.neon.tech/t/compute-time-usage-calculation/649/2).

The downside is that there can be an initial [cold server startup time](https://cloud.google.com/run/docs/about-instance-autoscaling#idle-instance) for **Google Cloud Run**.