Usage & Limits

The AI assistants run on your team’s own Anthropic API key. A team Owner adds the key once — see Using your own API key — and from then on every model call the assistants make is billed to your own Anthropic organization.

Instruqt does not spend tokens on your behalf. AI usage is not included in your Instruqt subscription.

Without a key

Until your team has an active key, the AI features are switched off. The lab assistant, Create with AI, and research all show a blocking message instead of starting work. Team Owners are sent to the API keys page; everyone else is asked to speak to an Owner.

Research documents you already have stay editable by hand. Only the AI parts — the chat and the create actions — are blocked.

The rest of Instruqt is unaffected. Editing, publishing, and running labs all work as normal.

What counts as usage

Every message you send an assistant spends tokens, and so does the work it does on its own between your messages, which is most of it. Some pages start AI work as soon as you open them — a research profile that isn’t published yet, or a lab whose plan is still a draft — so they spend tokens even if you don’t type anything.

Work done in your team by an Instruqt employee, such as onboarding or a support session, is never billed to you.

Where to see your spend

Instruqt does not show a spend figure. Your usage is visible in the Anthropic console for the organization that owns the key.

What stops a run

The limits that apply are your Anthropic organization’s, not Instruqt’s:

Out of credit or a revoked key — the run fails.
Rate limits — set by your Anthropic tier. Raise them with Anthropic.
Removing the key — stops runs already in flight.

In each case the turn fails and the chat tells you why. Everything already written is kept, so you can fix the key and start again.