Neon Cost Performance

Compute ran continuously for 30 days (no scale-to-zero)

Part of the Cost & ops health check · fix arrives as a guide

What it is

A compute endpoint has run continuously for the last 30 days without ever scaling to zero.

Why it matters

Neon's pricing assumes compute suspends when idle. Something is holding it awake around the clock — usually a health check or a background poll — so you are paying always-on prices for a database that could be idle most of the day.

How to fix it

Find what is connecting during quiet hours: uptime monitors, cron jobs, and long-lived pools with no idle timeout are the usual causes. Confirm autosuspend is enabled and set to a short window.

Run them all on your app

Connect your repo and your live services with read-only scopes. The first scan is free, and nothing changes without your approval.