---

name: cost-gcp description: Estimate GCP infrastructure costs for any stack using live pricing data. Use when user asks about Google Cloud pricing, cost estimation for Cloud Run, GKE, Cloud SQL, Compute Engine, or any GCP service combination.

GCP Cost & Ops Evaluator

Estimate costs and operational overhead for GCP-based infrastructure using real-time pricing data.

Before Answering GCP Cost Questions

IMPORTANT: Always verify tool availability before providing cost estimates.

Step 1: Check Available Tools

Check if the following tools are available:

GCP MCP Tools (preferred):

• mcp__gcp-mcp__search_documentation - Search GCP docs
• mcp__gcp-mcp__read_documentation - Read specific pricing pages

gcloud CLI (alternative):

• gcloud billing - Billing and pricing commands
• gcloud services - Service information

Pricing API:

• Cloud Billing Catalog API

Step 2: If Tools Are Unavailable

If GCP MCP or CLI tools are not accessible, help the user set them up:

Option A: GCP MCP Server Requirements: Python/uvx + GCP credentials

Copy# Install GCP MCP (if available)
uvx mcp-server-gcp

# Or use the documentation MCP
uvx mcp-server-gcp-docs

Option B: gcloud CLI

Copy# Install gcloud CLI
# macOS
brew install google-cloud-sdk

# Or download from https://cloud.google.com/sdk/docs/install

# Authenticate
gcloud auth login
gcloud config set project YOUR_PROJECT_ID

Option C: Enable Billing API

Copy# Enable Cloud Billing API
gcloud services enable cloudbilling.googleapis.com

# List billing accounts
gcloud billing accounts list

Ask the user which option fits their environment before proceeding.

Step 3: Fallback to Web Search

If no tools are available and user cannot install them, use web search to fetch current pricing from:

• https://cloud.google.com/pricing
• GCP Pricing Calculator: https://cloud.google.com/products/calculator

How to Get Live Pricing

Using gcloud CLI

Copy# List available SKUs for a service
gcloud services list --available | grep -i compute

# Get billing information
gcloud billing accounts list

# Export billing to BigQuery for analysis
gcloud billing accounts describe ACCOUNT_ID

Using Cloud Billing Catalog API

Copy# List all services
curl -H "Authorization: Bearer $(gcloud auth print-access-token)" \
  "https://cloudbilling.googleapis.com/v1/services"

# List SKUs for Compute Engine
curl -H "Authorization: Bearer $(gcloud auth print-access-token)" \
  "https://cloudbilling.googleapis.com/v1/services/6F81-5844-456A/skus"

# Common Service IDs:
# Compute Engine: 6F81-5844-456A
# Cloud SQL: 9662-B51E-5089
# Cloud Run: 152E-C115-5142
# Cloud Storage: 95FF-2EF5-5EA1
# GKE: 6F81-5844-456A (uses Compute)

Using Pricing Calculator API

Copy# Use the GCP Pricing Calculator for estimates
# https://cloud.google.com/products/calculator

# Export estimates via URL sharing

Standalone Usage

Can be invoked directly for cost estimation:

• "How much would Cloud Run + Cloud SQL cost?"
• "Estimate GCP costs for a Python API"
• "Compare GKE Autopilot vs Standard pricing"

Cost Estimation Process

1. Identify services needed for the stack
2. Query live pricing using available tools
3. Factor in sustained use discounts (automatic)
4. Calculate baseline costs (minimum viable)
5. Estimate at scale (10x baseline load)
6. Identify cost drivers and optimization opportunities

Service Categories & Tools

Compute Services

Database Services

Storage & Networking

Output Contract

Copygcp_cost_estimate:
  description: "<what's being estimated>"
  region: "<GCP region>"
  pricing_source: "<mcp|cli|api|web|calculator>"
  pricing_date: "<when pricing was fetched>"

  components:
    - service: "<service name>"
      config: "<configuration>"
      monthly_cost: "<$X>"
      sustained_use_discount: "<X%>"
      notes: "<any notes>"

  baseline_monthly:
    compute: "<$X>"
    database: "<$X>"
    storage: "<$X>"
    networking: "<$X>"
    other: "<$X>"
    total: "<$X>"

  at_10x_scale:
    compute: "<$X>"
    database: "<$X>"
    networking: "<$X>"
    total: "<$X>"

  cost_drivers:
    - "<primary cost driver>"

  discounts_applied:
    - type: "<sustained-use|committed-use|free-tier>"
      amount: "<$X or X%>"

  savings_opportunities:
    - type: "<committed-use|preemptible|spot>"
      potential_savings: "<X%>"
      commitment: "<1yr|3yr>"

  ops_overhead: "<low|medium|high>"
  required_gcp_knowledge: "<basic|intermediate|advanced>"

  optimization_tips:
    - "<tip 1>"
    - "<tip 2>"

Cost Optimization Strategies

Query these using GCP tools:

Committed Use Discounts

Copy# List committed use discounts
gcloud compute commitments list

# Create a commitment (save 57% for 1yr, 70% for 3yr)
gcloud compute commitments create my-commitment \
  --region=us-central1 \
  --resources=vcpu=4,memory=16GB \
  --plan=twelve-month

Preemptible/Spot VMs

Copy# Create preemptible VM (up to 80% savings)
gcloud compute instances create my-vm \
  --preemptible \
  --machine-type=e2-medium

# Or use Spot VMs
gcloud compute instances create my-vm \
  --provisioning-model=SPOT \
  --machine-type=e2-medium

Recommender API

Copy# Get cost recommendations
gcloud recommender recommendations list \
  --project=PROJECT_ID \
  --location=global \
  --recommender=google.compute.instance.MachineTypeRecommender

# Get idle resource recommendations
gcloud recommender recommendations list \
  --recommender=google.compute.instance.IdleResourceRecommender

Budget Alerts

Copy# Create budget alert
gcloud billing budgets create \
  --billing-account=ACCOUNT_ID \
  --display-name="Monthly Budget" \
  --budget-amount=1000USD \
  --threshold-rule=percent=50 \
  --threshold-rule=percent=90

GCP-Specific Pricing Notes

• Sustained use discounts: Automatic 30% off for consistent monthly usage
• Committed use: 57% (1yr) or 70% (3yr) savings
• Preemptible/Spot: Up to 80% off for fault-tolerant workloads
• Free tier: Many services have always-free tier
• Per-second billing: Most services bill per-second (1 min minimum)
• Network egress: Free within same zone, charges between zones/regions

GCP vs AWS Cost Comparison Tips

When comparing to AWS:

• Cloud Run often cheaper than App Runner for variable traffic
• Cloud SQL slightly cheaper than RDS at baseline
• GCP sustained use discounts are automatic (AWS requires RIs)
• Secret Manager is significantly cheaper than AWS Secrets Manager
• Check data transfer costs - can vary significantly