Whose API keys does the ExpressionEngine add-on use?
Yours. You paste an OpenAI, Gemini or Perplexity key into the settings screen, it is stored in your own database, and it is sent to nothing except the provider it belongs to. There is no account with us and no server of ours in the path.
What does a run cost on ExpressionEngine?
One run asks each active question to each selected engine once. Ten questions across three engines is thirty API calls, which is cents at current provider pricing. Inside the reuse window a repeated run makes no calls at all.
Why is the Shopify app a subscription and the EE add-on is not?
Because they are different products. The Shopify app runs on our servers and measures at product level, so there is a real recurring cost behind it. The ExpressionEngine add-on is self contained and runs on your own keys, so there is nothing recurring to charge for.
What happens if one engine fails?
That measurement is recorded as missing and the run continues. Engines are never substituted for one another. A number you cannot trust is worse than a gap you can see.
Which ExpressionEngine versions are supported?
EE 5, 6 and 7, from a single package. Scheduled runs use the CLI on EE 6 and 7, and a token protected URL on EE 5, which has no CLI.