Cloud Serverless Functions Cost Comparison Calculator

Enter one shared workload and up to three provider plans. The shared workload includes monthly requests, average execution time, allocated memory, vCPU, temporary storage, warm-request share, warm instances, warm hours, outbound transfer, logging, and a monthly budget.

Each plan can use its own provider, region, service plan, billing mode, architecture, request price, free grant, compute price, warm-capacity price, storage price, transfer price, logging price, discount, fixed monthly cost, and migration cost.

Configured plans are ranked by monthly planning cost. The result also shows first-year cost, cost per million requests, cost per execution, estimated average concurrency, budget position, and the possible saving against the lowest configured plan.

AWS Lambda and Azure Functions consumption plans commonly use request charges plus memory-based execution measured in GB-seconds. Google Cloud Run functions operate as Cloud Run services and can use request-based CPU and memory billing.

Select memory-only execution billing when the provider charges requests and GB-seconds. Select separate CPU and memory billing when vCPU-seconds and GB-seconds are priced independently. A complete monthly plan option supports bundled or negotiated quotes.

The calculator uses the exact duration, memory, and vCPU workload entered. It does not assume that similarly sized function configurations have equal performance.

AWS Lambda Provisioned Concurrency charges for configured warm capacity and uses a separate execution price while that capacity serves requests. Azure Functions Flex Consumption can charge an Always Ready baseline, Always Ready execution time, and executions. Cloud Run functions can use minimum instances or instance-based designs that create an idle or baseline cost.

Enter the percentage of requests expected to use warm capacity, the number of warm instances, monthly warm hours, and the memory and vCPU allocated to each warm instance. Each provider plan can then use separate warm baseline and warm execution rates.

If warm request or warm execution pricing is left blank, the calculator falls back to the plan's normal request or execution rate. This keeps the comparison usable when the provider uses the same rate for both paths.

Free requests and free compute grants are entered per provider plan and applied to on-demand usage. This avoids applying consumption-plan grants to provisioned, always-ready, or premium capacity when the provider does not allow it.

Temporary or ephemeral storage is calculated only above the included amount entered for the plan. Outbound data transfer and logging ingestion are calculated separately because they can materially exceed the function execution charge.

Trigger delivery, API gateways, storage accounts, build services, artifact registries, queues, databases, and private networking are not automatically included. Add them under fixed monthly services when they belong in the comparison.

Product names, hosting plans, billing structures, free grant rules, and warm-capacity models were checked on June 25, 2026. Prices remain blank because actual rates depend on region, architecture, plan, runtime, commitment, currency, free-tier eligibility, and account agreement.

Use each provider pricing page or account calculator to enter the current effective rates for the exact configuration. This prevents one regional example from being presented as a universal price.

• Compare AWS Lambda, Azure Functions, and Google Cloud Run functions using one workload.
• Compare request-plus-GB-second and CPU-plus-memory billing.
• Measure provisioned concurrency, always-ready, and minimum-instance cost.
• Apply provider-specific free requests and compute grants.
• Add temporary storage, transfer, and logging charges.
• Compare architecture or commitment discounts.
• Add platform, gateway, build, storage, or support costs.
• Measure monthly operating and first-year migration cost.
• Check each serverless plan against a monthly budget.

The result does not automatically include taxes, premium cloud support, API gateways, Eventarc, Event Grid, SNS, SQS, storage accounts, Cloud Build, Artifact Registry, container registries, queues, databases, private networking, public IPv4 addresses, migration labour, negotiated credits, or services not entered.

Similar memory and vCPU allocations do not guarantee similar function performance. Runtime startup, processor architecture, concurrency, cold starts, timeout limits, network latency, event delivery, and managed integrations differ. Test a representative workload before choosing a provider.