You bring your own AI key and pay your provider directly, so what is this costing me used to mean opening their dashboard. The AI Usage card at the bottom of the Connect AI tab answers it in the plugin.
The four figures:
- Total Jobs: every AI call this plugin has ever made
- Last 30 Days: jobs in the last 30 days
- Spend, 30 Days: your real token counts priced at your provider's published rates
- Errors, 30 Days: failed calls, which cost you nothing, shown so that a quietly broken key cannot hide
These are a close estimate, not an invoice. They are your provider's own token counts at their published prices, which is as accurate as anybody can be without billing you twice.
Recent AI Jobs:
The last ten calls, each with what it was for, whether it completed, which provider and model ran it, what it cost, how long it took and when it happened. A failed job prints the provider's own error message underneath it, which is usually the fastest way to see that a key has expired or a model has been retired.
The monthly soft budget:
Set Monthly Soft Budget in the Routing card and this screen shows a progress bar and warns you when you cross it.
It is a reminder, not a lock. Nothing stops working when you go past it and nothing is ever blocked mid build. Set it to 0 to turn the warning off entirely.
Real spending limits belong in your AI provider's account, where they can actually be enforced. A budget in a plugin can only ever be advice.
What a funnel actually costs:
Roughly one to three cents per page on a recommended model, paid to your provider, or nothing at all on the Gemini free tier. Rewriting a page or a section costs about the same as building it. Restyling a funnel with a different theme costs nothing at all, because no AI call is involved.
Voice dictation costs nothing either. Your browser does the listening, so it never touches your key, your usage or your budget.