Botperch

Cost calculator

What would it cost each month?

Describe the machine your agent needs and how it runs. Every provider is priced from its own published rates, and each result opens to show the arithmetic, the plan it assumes and anything that stops it running as asked.

27 providersList prices, with the arithmeticRuns in your browser

Start from

The machine
How it runs
How busy it is

Only providers that bill for CPU actually used look at this. An agent waiting for messages is near 0%; a build or a test run is near 100%.

For providers that put an idle machine to sleep and wake it on a request: the share of its hours it has to be awake. An agent that only answers webhooks can sleep; one that holds a chat connection open cannot.

The AI model

These two numbers are our assumptions, not measurements. An agent re-sends its instructions and history on every call to the model, and one message can take several calls, so 100,000 tokens or more per message is normal. Change them to match your agent.

Per-token prices are the median of the model hosts that keep no data, from TokenWatch, read 2026-10-31. Plan allowances are estimates from Real API Pricing. How the model line works.

    Machine prices per provider: list prices only, excluding the model, egressData the machine sends out to the internet, such as downloads to a user or uploads to another service. Some providers charge per GB; most agents send very little., taxes and anything a provider does not publish. How the numbers are made.

    Reading the result

    Runs as asked means the provider sells this size, keeps the files if you asked for that, and has no session limit shorter than your sessions. Needs a change means the figure is for the nearest thing it sells, and the reason is listed when you open the row.

    The cheapest number is rarely the whole answer. A provider that bills only for CPU actually used is cheap for an agent that mostly waits and expensive for one that compiles all day; a flat monthly plan is the reverse. Move the two sliders to see which providers are sensitive to them.

    Wondering whether to rent a plain server instead? Server or sandbox puts every VPS here against every sandbox, by hours awake a day.

    Wondering which half of the bill matters? Model or machine puts the machine beside the model for every ready-made workload.

    To compare two or three providers on features and cost together, use side by side. To start from what you want to run instead of numbers, see the use cases.