A. Supply charges
The seats you pay for, used or not.
Tick what you're on and pick the tier. The price follows the tier, not the logo.
B. Usage charges
One workload, every model's price.
Set your monthly token throughput once, then read each model's bill for that same workload. Tick the ones you actually run and they move onto the statement.
Meter reading
tokens this period
| Include | Model | In / M | Out / M | Relative cost | Per month |
|---|
C. Amount due
Your total, monthly in AUD.
Total this month
Nothing selected yet.
Where it goes
These figures are estimates
This calculator gives indicative estimates only. Figures are based on published list prices as at September 2026 and the inputs you choose. Actual charges depend on each vendor's current pricing, plan terms, exchange rates, taxes, fees and your real usage, and can change without notice. Nothing here is financial, tax or purchasing advice. Check current pricing with each vendor before you commit to a plan or budget.
Fine print
Prices are US list rates published in early September 2026, converted at the rate in the controls above. Change it to whatever your card actually gave you. Some local storefronts price in AUD directly, so treat the AUD figures as a close estimate, not a quote.
GST: most of these vendors charge 10% GST to Australian consumers. If you've supplied a valid ABN, some will zero-rate you, so untick it. The 3% card fee is the typical international transaction charge on an Australian card, off by default because plenty of cards no longer charge it.
API costs assume standard, non-batch processing on short-context requests. Batch runs are roughly half price, prompt caching cuts cached input by about 90%, and long-context requests on several models bill higher. Read these as a sensible upper-middle estimate. Per-seat plans are costed as one seat.
Sources: Claude pricing, Claude API pricing, ChatGPT pricing, OpenAI API pricing, Gemini API pricing, xAI models, AI subscription price index. Nothing you enter leaves your browser.