> ## Documentation Index
> Fetch the complete documentation index at: https://elephas.app/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# How AI Usage and Credits Work

> How the Elephas AI usage in your plan works: what uses it, what never does, where to check it, when it resets, and how to keep asking when it runs out.

<Warning>
  Draft. This article depends on an open product question and is not ready to publish.
</Warning>

Every paid plan includes a monthly amount of Elephas AI usage. Only answers from Elephas AI use it. Indexing your documents never does, and neither does asking with your own API key or Local AI.

## Before you start

* You need an active trial or plan. See [plans and pricing](/docs/plans-and-billing/plans-and-pricing).
* Usage applies to Elephas AI, the option labelled **Elephas AI (included with your plan)** in **Settings › AI**.

## Check your usage

<Steps>
  <Step title="Open Settings › AI">
    On Mac, choose **Elephas › Settings…** or press **⌘,**. On Windows, choose **File › Settings…** or press **Ctrl+,**. Then choose **AI**.
  </Step>

  <Step title="Read the meter">
    On a paid plan, the meter is labelled **Included AI usage** and reads, for example, "Included AI usage: 40% used this month. Resets" followed by the reset date. On the free trial it is labelled **Trial AI usage** and shows when the trial ends.
  </Step>

  <Step title="Refresh if needed">
    Click **Refresh Usage** to fetch the latest numbers. If Elephas can't reach its server, it shows the last known figures with "Couldn't refresh. Last known at" and the time.
  </Step>
</Steps>

## How it works

### What uses your allowance

* **Answers from Elephas AI.** Each question uses some allowance. Longer questions, more passages and longer answers use more.
* **Agent steps.** When Elephas takes several steps to answer, such as reading whole sources, each step counts. Before a question that reads whole sources, Elephas warns: "This question reads whole sources, so it may use noticeably more of your AI allowance than usual."

### What never uses it

* **Indexing.** Reading your files, OCR and building the search index run on your computer. Always.
* **Search and citations.** Finding passages and opening sources are local.
* **Your own API key.** Your provider bills you directly. See [use your own API key](/docs/ai-options/use-your-own-api-key).
* **Local AI.** The model runs on your computer. See [run Local AI](/docs/ai-options/run-local-ai).

### Credits

Plan cards show the monthly amount as AI credits, for example "Includes 3,000 AI credits of usage each month." Credits are a display unit for your included usage. They are not money, they cannot be bought or refunded, and they have no cash value.

### When it resets

Your included usage resets once a month. The reset date is shown under the meter. On an annual plan you still get a monthly amount, not twelve months at once. Unused usage does not roll over.

The free trial is different: its usage is one amount for the whole 7 days, with a daily share. See [free trial](/docs/plans-and-billing/free-trial).

### When you get close or run out

* At about 80%, Elephas shows a notice so you are not surprised.
* When it runs out, Elephas AI pauses and shows "Your Elephas AI allowance is used up until" followed by the reset date. You are never charged extra.
* You can keep asking right away by switching to your own API key or Local AI in **Settings › AI**. Both are included with Personal and Professional.
* If you regularly run out, Professional includes more usage each month. See [upgrade your plan](/docs/plans-and-billing/upgrade-your-plan).

### On a Team plan

A Team shares one pool of Elephas AI usage, sized by its paid seats. The Team Owner or an Admin can set a cap for each member. If you reach it, you see "You've reached the limit your Team set for you." Team workspaces cannot use your own API key. See [team seats and billing](/docs/teams/team-seats-and-billing).

## Troubleshooting

<AccordionGroup>
  <Accordion title="Usage isn't available yet.">
    Elephas has not received your usage figures yet. Check that you are signed in and online, then click **Refresh Usage**.
  </Accordion>

  <Accordion title="Elephas AI allowance used up. Some requests are still settling">
    A recent answer is still being counted. Wait a moment and try again.
  </Accordion>

  <Accordion title="My allowance ran out in the middle of an answer">
    Elephas stops and shows "Your AI allowance ran out during this answer. Check usage in Settings." See [AI allowance used up](/docs/troubleshooting/ai-allowance-used-up).
  </Accordion>
</AccordionGroup>

## Related articles

<CardGroup cols={2}>
  <Card title="Elephas AI" href="/docs/ai-options/elephas-ai" />

  <Card title="AI allowance used up" href="/docs/troubleshooting/ai-allowance-used-up" />

  <Card title="Choose how Elephas answers" href="/docs/ai-options/choose-how-elephas-answers" />

  <Card title="Plans and pricing" href="/docs/plans-and-billing/plans-and-pricing" />
</CardGroup>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.