Claude Plan Calculator

How to use it

Pick the usage profile closest to how you work, or choose Custom and type your own numbers. Set how many days a month you use Claude Code, how many sessions you run each day, and roughly how many tokens a session uses. Then say how your work splits between the four models. The result updates as you type.

The big number is what your usage would cost on pay-as-you-go API billing. The table underneath sets that number next to the monthly price of Pro, Max 5x and Max 20x. If you already use Claude Code, run npx ccusage@latest to see your real token counts and enter those instead of our rough profile numbers.

How it works, and what it cannot tell you

We multiply your tokens by the published API price of each model. That gives an "API equivalent" cost: what the same work would cost if you paid per token. We then compare it with the flat monthly price of each subscription plan.

What we cannot do is tell you whether a plan will cover your usage. Anthropic does not publish exact Claude Code limits for each plan. In general, the heavier plans give you more usage in each 5-hour window and each week, but the exact amounts can change. For the real limits, read the Claude pricing page and Anthropic's help centre. Our profile numbers and the default plan prices are rough examples, not measurements, so edit them. Agentsforce is an independent project and is not affiliated with Anthropic.

FAQ

Will Pro or Max cover my usage?

This tool cannot say. It only shows what your usage is worth at API prices. If that figure is well above a plan's price, the plan is probably good value, provided your usage fits inside its limits. If you keep hitting the limits, the next plan up gives more room.

Why are cached reads such a big part of the usage?

Claude Code sends your conversation and project context again on every step. Cached reads are billed at a small fraction of the normal input price, but there are so many of them that they often make up most of the tokens in a session.

Does a subscription include anything the API does not?

Yes. A subscription also covers the Claude chat apps, which the API does not. This calculator only compares Claude Code usage, so it ignores that benefit.

Where do I find my real token numbers?

Run npx ccusage@latest in your terminal. It reads the local logs Claude Code writes and shows tokens and cost by day. You can then turn those stats into a card with our Claude Code Wrapped tool.

Are the API prices up to date?

We checked them in October 2026. They are USD per million tokens, for prompts up to 100K tokens. Prices change, so confirm on Anthropic's pricing page before you decide.

Once you know your plan, Agentsforce ships 38 ready-made specialists that make each Claude Code session more useful.