AI and usage

Building your brain and answering questions both use AI models. Here is who pays for that, and how to see what your workspace has used.

Early access includes AI usage

A new workspace starts with an AI allowance on Noetix's own key, so you can build your brain and ask questions without a key of your own. Settings › Usage shows how much of it is used, as a percentage.

While you're on the allowance, the chat runs on Auto, the model Noetix picks.

Bring your own AI key

When the allowance is used up, building and asking need your own key. An admin adds it in Settings › AI provider, from OpenRouter, OpenAI or any OpenAI-compatible provider.

• From then on, the AI work runs on your key and is billed by your provider.
• If a build stopped when the allowance ran out, adding a key builds what was left.
• Noetix never shows your key again after you save it: only the provider and its last four characters.

Choosing models

With your own key, the model menu in the chat opens up. An admin chooses which models appear there in Settings › AI provider. Building the brain always uses the same model, so the brain stays consistent whichever model people chat with.

Your plan's limits

Settings › Usage shows, for everyone in the workspace, what the plan includes and how much is used: stored items, pages, skills, members, connected sources, and API and MCP calls this month. Your website doesn't take one of the connected source slots.

When something would go over a limit, Noetix says which limit and what to do, instead of failing quietly. Editing never counts against a limit; only adding more does.

API and MCP calls are warned about once they reach the month's limit, and stopped a little past it, so an agent mid-task isn't cut off the moment you reach it.