Usage-based pricing calculator

Pricing an AI feature per token, or an API per event? Plug in your usage and costs to see monthly revenue, cost, gross margin, the bill each customer gets and your break-even price. Two modes: LLM tokens (per-1M-token or flat fee + overage) and generic events (free allowance + graduated tiers). Any currency. It runs in your browser; nothing is sent.

Model and infrastructure prices change; check your providers' current pricing. Defaults are examples only.

Next step: meter it

A price only works if usage is counted exactly once. Start on the free tier (100k events, 3 meters), drop in the LLM metering helper for OpenAI / Anthropic / Vercel AI SDK, or validate an event without a key in the playground.

Embed this calculator

Free to put on your blog, docs or pricing page. Copy the snippet:

Copy embed code · Preview the embed

FAQ

How do I price LLM tokens for my customers?

Start from your model cost per 1M tokens, then add your margin. The break-even price is your blended model cost per 1M tokens; most AI SaaS teams charge several times that to cover failed requests, support and infrastructure.

Flat fee plus overage or pure per-token?

Flat fee plus overage gives predictable revenue and a simple plan page; pure per-token tracks cost exactly. The calculator shows the break-even overage price once the flat fee is used up.

How do tiers and free allowances work here?

Usage above the free allowance is billed at the tier 1 price up to the tier 1 limit, then at the tier 2 price (graduated tiers). Set the tier limit to 0 for a single flat rate.

Is my data sent anywhere?

No. The calculator runs entirely in your browser; nothing is sent or stored.

How do I actually meter this usage?

Send each usage event to a meter with an idempotency key so retries never double-bill, then read the monthly rollup per customer. Usage Lab does that on a free tier.