What this guide covers
- Which action actually uses a credit
- When a credit is deducted and when it is not
- What happens if you reject a generated result
- How insufficient credits, cooldowns, and limits affect generation
This guide explains the credit rules around the AI Adjustment Tool. Credits are tied to real generation requests. Opening the tool, choosing a student, reviewing a preview, or saving an accepted result does not create a new charge on its own.
Step 1: Know what actually uses a credit
A credit is used when Superadjust runs a generation request and returns a result for review.
- Opening Adjust Resource, selecting a student, choosing a subject, or filling in the side panel does not use a credit by itself.
- The clearest rule is simple: no generation, no credit used.
Quick credit map
| Action | Rule |
|---|---|
Open Adjust Resource | No credit used. Browsing, selecting a student, or entering inputs does not deduct a credit. |
Generate a resource | 1 credit used. The charge is tied to the generation request. |
Review the preview | No extra credit. Reading the review modal does not create a second charge. |
Accept and save | No extra credit. Saving to the student record or also saving as evidence does not use another credit. |
Reject after review | Credit already used. The completed generation still counts as usage. |
Step 2: Check the balance and button state before you run it
The Generate button shows the cost before the request starts, for example "Generate - 1 credit". If no student is selected, the button stays disabled. That is a workflow guardrail, not a credit event.
Before you generate, check the credit counter in the navigation and make sure your setup is complete so you do not spend a credit on a preventable retry.
Credit flow
Credit flow
Step 3: Understand when the credit is deducted
The product documentation is consistent on the main user rule: credits are tied to the generation event, not to the later save step. That means the credit is used when you run the request and Superadjust creates the preview for review.
- If you reject a preview and regenerate, that new generation uses another credit.
- Rejecting a result does not refund the credit. The AI call has already run, so the generated preview still counts as usage.
Important: Rejecting a result does not refund the credit. The AI call has already run, so the generated preview still counts as usage.
Step 4: Know the limits that protect credit usage
Superadjust includes several guardrails to prevent accidental overuse:
- Only one concurrent generation is allowed per teacher. If a request is already in flight, the button is disabled so you cannot accidentally spend credits twice on the same moment of work.
- There is also a minimum 5-second cooldown between requests to prevent accidental double-generation.
- If you do not have enough credits, generation is blocked before the request runs and the product shows the purchase or top-up flow.
- If a preview expires after 30 minutes, you cannot save it later. You need to generate again, which means using another credit.
Why this matters
Credits make AI usage predictable for schools and stop accidental overuse. Just as importantly, the rules keep the workflow honest. A generation only becomes evidence when a teacher reviews it and confirms the save, but the cost sits with the generation event itself.
Common mistake
Assuming the credit is only used if you accept the output. In the AI Adjustment Tool, the credit is spent when the generation runs, so it is worth checking your student, subject, and prompt before you click Generate.
What to do next
Now that you know the credit rules, the next step is to use Adjust Resource with cleaner inputs and fewer retries. Start with one student, review the preview carefully, and only regenerate when the teaching goal genuinely needs a new result.
Next guide
Inviting colleagues →
Send school-linked invitations, earn credits when colleagues join, and track invite status.