Compare costs across virtual machines, storage, managed databases, Kubernetes, serverless functions, NAT, load balancing, and other infrastructure that contributes to a cloud bill.
The visible hourly or monthly rate is only one part of a useful estimate. Resource choices, availability requirements, data movement, performance settings, and purchase models can all change the amount you actually pay.
VM cost can change with instance size, operating system, runtime, region, commitments, Spot or other interruptible capacity, attached storage, public IPs, and related network services. A discounted commitment should not be compared with another provider's on-demand rate as though the terms were equivalent.
Managed PostgreSQL, MySQL, Redis, and SQL Server costs can extend beyond compute. High availability, replicas, storage, IOPS, backups, licensing, support, and transfer can materially change the result.
Object and block storage can be affected by storage tier, requests, retrievals, snapshots, IOPS, throughput, lifecycle transitions, replication, and data leaving the service. The cheapest price per GB is not always the cheapest workload.
NAT gateways, load balancers, public IPs, processed traffic, cross-service transfer, internet egress, logging, and security features can sit outside the headline compute price. They are easy to miss when an estimate focuses only on servers.
Kubernetes and serverless platforms may combine control-plane fees, worker or pod compute, requests, execution time, memory, vCPU, storage, networking, logging, and always-on capacity. The architecture decides which of those costs matter.
Choose the tool that matches the resource or architecture you want to compare. Each page focuses on the billing inputs and assumptions that matter for that service.
Compare AWS EC2, Azure Virtual Machines, Google Compute Engine, or custom VM plans using runtime, commitments, Spot usage, storage, bandwidth, IPs, load balancers, and first-year cost.
Compare Amazon S3, Azure Blob Storage, Google Cloud Storage, or custom object storage plans using storage tiers, requests, retrievals, egress, lifecycle transitions, replication, and migration cost.
Compare Amazon RDS for PostgreSQL, Azure Database for PostgreSQL Flexible Server, and Google Cloud SQL using compute, HA, replicas, storage, IOPS, backups, transfer, support, and migration cost.
Compare Amazon RDS for MySQL, Azure Database for MySQL Flexible Server, and Google Cloud SQL using compute, HA, replicas, storage, IOPS, backups, transfer, extended support, and migration cost.
Compare Amazon ElastiCache, Azure Managed Redis, and Google Cloud Memorystore using node, serverless, memory, request, backup, transfer, support, and migration costs.
Compare Amazon RDS for SQL Server, Azure SQL Managed Instance, and Google Cloud SQL using compute, licensing, HA, storage, IOPS, backups, transfer, discounts, and migration cost.
Compare Amazon EKS, Azure Kubernetes Service, and Google Kubernetes Engine using cluster fees, worker or pod compute, storage, load balancers, NAT, transfer, logging, backups, discounts, and migration cost.
Compare AWS Lambda, Azure Functions, and Google Cloud Run functions using requests, execution time, memory, vCPU, warm capacity, storage, transfer, logging, discounts, and migration cost.
Compare AWS EBS gp3, Azure Premium SSD v2, Google Hyperdisk Balanced, and custom block storage costs using capacity, IOPS, throughput, snapshots, migration, and budget.
Compare AWS NAT Gateway, Azure NAT Gateway, Google Cloud NAT, and custom managed NAT costs using gateway hours, protected instances, processed traffic, public IPs, transfer, logging, and migration cost.
Compare AWS, Azure, Google Cloud, and Cloudflare load balancer costs with editable hourly, capacity, traffic, WAF, logging, transfer, and setup inputs.
A provider comparison is only useful when the workloads are genuinely comparable. Keep the region, runtime, availability target, performance requirement, storage amount, backup policy, transfer pattern, and support assumptions as close as possible.
Purchase models also need to be compared on equal terms. On-demand, reserved or committed capacity, and Spot or interruptible capacity trade flexibility, commitment, and interruption risk differently. A lower rate can come with a condition that changes whether the option fits the workload.
First-year cost can also include migration, setup, data transfer, support, or parallel-running expenses that do not appear in a simple steady-state monthly rate. Those costs matter when the decision is about moving an existing workload rather than starting from zero.
Traffic leaving a service, region, or cloud can add charges that are not visible in the resource's base price.
Resilience and recovery features can add storage, compute, or transfer costs even when the primary workload stays the same.
Gateways, load balancers, reserved capacity, warm instances, and other resources may continue costing money when application traffic is low.
Database licenses, extended support, monitoring, logging, and security services can materially change a production estimate.
Beeija calculations are planning estimates, not provider invoices. Cloud pricing can vary by provider, region, resource type, purchase model, currency, discounts, taxes, support plan, and actual usage.
When provider pricing is built into a tool, the aim is to make the source, checked date, assumptions, and editable inputs clear enough that you can see what is driving the comparison and update rates when needed.
An AI product can have two separate cost layers: model or API usage, and the cloud infrastructure running the rest of the application. Use the AI category when you need to estimate that model-side cost.
AI Cost Calculators