Google Cloud Cost Optimization
OhChimp cuts your Google Cloud bill: rightsize Compute Engine, move pd-ssd disks to pd-balanced, and catch idle Cloud SQL, with each fix written as a plan.
What OhChimp optimizes for Google Cloud
Some of the cost levers OhChimp checks for Google Cloud. Each one becomes a reviewable plan you approve before anything changes.
Idle and oversized Compute Engine
OhChimp pulls Google Cloud Recommender machine-type recommendations to catch VMs sized larger than their real load, and flags stopped instances still billing for their disks. Each one becomes a resize or cleanup plan with the dollar delta and rollback steps.
pd-ssd disks that should be pd-balanced
Attached pd-ssd disks often run general workloads that never need SSD-tier IOPS, and pd-balanced lists roughly 41% cheaper per GB-month. OhChimp groups your attached pd-ssd disks and writes a non-destructive migration plan that snapshots and creates a parallel pd-balanced disk first, so the original stays intact until you verify the swap.
Unattached disks and unused static IPs
Persistent disks left after their instance is gone, and static IP addresses that show as not in use, keep billing every month for nothing. OhChimp finds both and drafts the release or delete plan, with the monthly figure attached.
Idle or oversized Cloud SQL
Cloud SQL instances sitting in a stopped or suspended state still cost money against their tier. OhChimp flags them and proposes deleting or resizing the instance, so you review the change before anything touches the database.
Committed Use Discount gaps
Steady Compute Engine usage running fully on-demand leaves money on the table. OhChimp groups your running machine types, reads your existing commitments, and shows where a one-year (around 37%) or three-year (around 56%) Committed Use Discount would pay off on the spend you already run.
GKE, Cloud Run, and Vertex AI waste
OhChimp reviews GKE node pools for autoscaling and rightsizing, estimates Cloud Run min-instance keep-alive cost, and flags services with zero request traffic. It checks Vertex AI for frontier-model token spend plus dedicated endpoints holding replicas while serving no predictions. It also reads your Terraform to map google_ resources and how they connect, so cost findings tie back to the code that provisions them.
Worked example
For example, say OhChimp finds a fleet of attached pd-ssd disks on low-throughput VMs, two static IPs flagged as not in use, and a Cloud SQL instance sitting in a stopped state. It bundles the disk migrations, the IP release, and the Cloud SQL change into reviewable plans, each with a confidence score and rollback steps.
You press apply, and OhChimp swaps in the cheaper pd-balanced disks, releases the IPs, and resizes Cloud SQL. OhChimp cannot yet measure what a single Google Cloud plan saves, so it does not mark one VERIFIED. The projected savings stay estimates, and when the 30 days of monitoring end the result reads "Inconclusive."
Google Cloud cost optimization FAQ
How does OhChimp connect to Google Cloud?
You grant a read-only service account key. OhChimp verifies it by reading your project, then pulls cost and configuration metadata through the Compute Engine, Cloud SQL, Recommender, and Cloud Monitoring APIs. It never stores application data, secrets, or workload contents.
Does OhChimp need write access or see my data?
OhChimp reads resource configs, billing data, and aggregated metrics with read-only access. It never stores your application data, secrets, or workload contents. The only writes happen when you apply a plan, and you are the one who clicks apply.
What Google Cloud costs can OhChimp reduce?
Idle and oversized Compute Engine VMs, stopped instances still billing for disks, pd-ssd disks that should move to pd-balanced, unattached disks and unused static IPs, idle or oversized Cloud SQL, low Committed Use Discount coverage, GKE node-pool rightsizing, idle Cloud Run min-instances, and Vertex AI token spend plus idle dedicated endpoints. It also reads your Terraform to map google_ resources and how they connect.
Who applies the changes, and can they be rolled back?
You do. Each fix is a reviewable plan with a confidence score, a risk level, and rollback steps. Nothing moves until you click apply, and OhChimp applies the matching code and infrastructure changes together. The pd-ssd migration plan is deliberately non-destructive: it builds a parallel pd-balanced disk and leaves the original until you confirm the swap.
How are the savings verified, and how is OhChimp priced?
Today Google Cloud savings stay estimates. OhChimp cannot yet measure what a single Google Cloud plan saves, so it does not mark one VERIFIED, and when the 30 days of monitoring end the result reads "Inconclusive." The only savings OhChimp measures today come from lowering the CPU and memory requests of a Kubernetes Deployment, StatefulSet, or DaemonSet. Pricing is flat monthly, OhChimp never takes a cut of your savings, and a year-one ROI guarantee lets you request a refund of your first 12 months of subscription fees if the verified savings on your Costs tab fall short of them. Estimated savings do not count.
The pd-ssd to pd-balanced lever, start to finish: GCP pd-ssd vs pd-balanced: cut disk costs 40%.
The Cloud Run min-instance keep-alive lever, priced end to end: Cloud Run min instances: idle is $9.72 or $49.25.
The unattached disk lever, including what the safety snapshot costs: GCP unattached disks: delete without paying twice.
Related integrations
Teams running Google Cloud usually run these too. OhChimp finds the waste in each and drafts every fix as a plan you approve.
Azure
Underused VMs, disk tiers, reservation coverage
AWS
Idle compute, storage class, reserved coverage
DigitalOcean
Oversized Droplets, Spaces storage, idle nodes