AI pricing and consumption
Every run of an AI application uses tokens: the text sent with the run (input tokens) and the text of the answer (output tokens). There is no plan to choose and no pack to buy: usage is priced per token, 1.00 € per million input tokens and 5.00 € per million output tokens, excluding VAT. The Pricing tab shows the two prices and where the week stands; the Consumption tab shows what your applications used. Both tabs are in Shop AI, under AI in the main menu.
Note — Your AI space is set up the first time you run an application. Nothing has to be configured before.
Check the price and the week so far
- Open Shop AI and click the Pricing tab.
- The Pricing card shows the price of Input tokens and of Output tokens, each per million tokens.
- The This week card shows the Input tokens and the Output tokens your runs used since Monday, and the Amount so far at that price.
Note — Your workspace is billed by the week, and the weekly invoice carries the tokens of that week. A run that fails uses nothing: it is not counted in the week's tokens. You can create as many applications and workflows as you need; only the tokens of their runs are counted.
See what your applications consumed
- Open Shop AI and click the Consumption tab. The figures cover This cycle by default.
- To look at another period, pick it at the top right of the Consumption card, then click Apply.
- Read the Runs, the Billed tokens, the Input tokens and the Output tokens, with the average per day.
- Below, the Applications card ranks what used the most Runs and Tokens, with the Total.
Note — Billed tokens are the input plus the output tokens of the runs that did not fail; the number of runs counts every run.