AI cost calculator

What would your AI cost to run?

Explore the API cost of your workflow. Change the volume and see where the money goes.

Describe your usage

Presets are examples to edit, not measured averages.

Include prompts, history, retrieved text and billed reasoning output. This calculator supports up to 100,000 input tokens per call.

Choose model rates

Standard text API pricing. No cache, batch or regional modifiers included.

Official provider pricing

Editing a rate switches to Custom rates. Models are listed for cost comparison; capability differs.

Estimated monthly API cost

— USD / month

Per task
—
Per year
—
Monthly input tokens
—
Monthly output tokens
—

Tasks × calls per task × (input tokens × input rate + output tokens × output rate) ÷ 1,000,000.

Text-model usage only. Implementation, hosting, storage, retrieval, tools, media, taxes and support need a separate estimate. No caching or batch discount is assumed.

Discuss the full workflow

What this estimate covers

This calculator estimates input and output token charges for text-model API calls. A task may involve several model calls; include all the calls the workflow needs, along with conversation history, retrieved documents and billed reasoning tokens.

The presets are illustrative workloads, not measured averages. Replace them with usage from your own tests when available. Model selection here compares rates, not suitability or answer quality.

What needs a separate estimate

Implementation, hosting, storage, retrieval, document parsing, paid tools, images, audio, taxes and support are outside this calculation. Prompt caching and batch discounts are not assumed. Use custom rates if your agreement or processing tier differs.

For a full project estimate, bring the workflow and the systems it needs to connect to.

Put the estimate in context.

Share the workflow, expected volume and integrations. We'll help define the implementation and operating scope.

Discuss the full workflow