Estimates only. Provider pricing changes often; check each provider’s current pricing page before budgeting.
Prices checked October 2026: standard (on-demand) API rates in USD per 1M tokens, from Anthropic, OpenAI, and Google’s official pricing pages. They are editable defaults, not live data – always check the provider’s current pricing page before budgeting a production workload.
About this tool
The AI API Cost Calculator helps developers, product managers, and startups estimate what they will actually spend when calling large language model APIs across the major providers. You pick a model preset for Claude, OpenAI GPT, or Gemini (or type your own per-million-token prices), plug in your expected token volumes, and the tool estimates the cost per run, per month, and per year, so you can budget before you build. It is built for anyone who bills by the token and needs real numbers, not guesses.
Reach for this tool when you are evaluating which AI provider fits your budget, scoping a new product feature, or auditing monthly API spend. It is especially useful when you are choosing between providers and want a quick apples-to-apples estimate before reading each provider’s full pricing page.
How to use it
- Pick a model from the preset list, or type your own input and output prices per 1M tokens.
- Enter the input tokens for one run of your workload.
- Enter the output tokens for one run.
- Enter how many runs you expect per month.
- Review the per-run, per-month, and per-year estimates.
- Switch presets or adjust token volumes to compare models before committing to a provider.
Pro tips
- Output tokens typically cost several times more than input tokens (about five times on many current models), so even modest response lengths can dominate your total bill.
- Long prompts can move a request into a higher price tier on some models (for example Claude Haiku 5.5 above 100K-token prompts, Gemini Pro above 200K, and OpenAI models above 272K input tokens); check the provider’s long-context rates.