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

# Usage

> Understand how queries, tokens, storage, compute, users, and tools are metered on Purna.

Your Purna account includes resource quotas that determine how much you can use each month. This page explains what each resource means, how it is measured, and what to expect.

## Overview

<CardGroup cols={3}>
  <Card title="Queries" icon="message">
    Daily new conversations
  </Card>

  <Card title="Tokens" icon="coins">
    Monthly conversation volume
  </Card>

  <Card title="Compute" icon="microchip">
    Code execution and job runtime
  </Card>

  <Card title="Storage" icon="hard-drive">
    Uploaded files and generated outputs
  </Card>

  <Card title="Users" icon="users">
    Team members on your account
  </Card>

  <Card title="Capabilities" icon="wand-magic-sparkles">
    Tool access controlled by plan
  </Card>
</CardGroup>

***

## Queries

Queries are the number of new chat conversations you can initiate per day. Continuing an existing conversation does not use another query.

Queries reset daily at midnight UTC.

<Tip>
  If you are running low on queries, continue working in an existing conversation instead of starting a new one. Follow-up messages in the same chat do not count as additional queries.
</Tip>

## Tokens

Tokens represent your total monthly chat volume. Every message you send to Purna and every response Purna sends back consumes tokens. Longer exchanges, large file context, tool results, and detailed outputs use more tokens.

A typical research query uses roughly **2,000 to 5,000 tokens**. Complex literature synthesis, file analysis, or long-form drafting can use more.

**What counts toward tokens:**

* Your messages and prompts
* Purna's responses, including formatted text, tables, and citations
* Context sent to and from AI tools during a conversation
* Summaries or extracted content from referenced files

<Note>
  Token usage resets at the start of each billing cycle. Monitor your current usage in account settings.
</Note>

## Compute

Compute measures how long private execution environments run for [Code Execution](/code-execution) and background jobs.

By default, code execution provisions a private server attached to your chat session. Background jobs use isolated cloud sandboxes for longer workflows.

**How compute is metered:**

* Active execution time is counted while code or jobs are running
* Paused or idle resources do not continue consuming runtime once shut down
* Long-running jobs and large analyses use more compute than quick transformations

<Tip>
  Keep related analysis in the same chat when possible. Follow-up code execution can reuse context and reduce repeated setup time.
</Tip>

## Storage

Uploaded files, generated artifacts, job outputs, images, and exported datasets count toward your storage quota. Storage is measured as total disk usage across your account.

Common storage contributors include:

| File type                | Typical size         |
| ------------------------ | -------------------- |
| CSV / TSV tables         | KB to hundreds of MB |
| H5AD or matrix files     | MB to multiple GB    |
| Images and figures       | KB to tens of MB     |
| Job outputs and archives | MB to multiple GB    |
| PDFs and documents       | KB to hundreds of MB |

<Note>
  Deleting a file removes it from your Files library and helps reclaim storage. Generated outputs attached to a job or chat may still be represented in the conversation history.
</Note>

## Users

Users are team members on your account. Each user gets their own login, chat history, private files, and access controlled by project and team settings.

User management is available in [Team settings](/guides/team). Invite new members, assign roles, and decide which work should be private or team-visible.

## Tools and capabilities

Tools and capabilities are controlled by your plan and your chat **Preferences**. Some capabilities are available to every account, while more resource-intensive tools may require a paid plan.

Examples include:

* Web and database search
* Code execution
* Background jobs
* Image and figure generation
* Deep Research
* Consilium
* Paper Writing

If you need a specific integration that is not currently available, contact [contact@purna.ai](mailto:contact@purna.ai) and we will work with you to evaluate it.

## Enterprise

Enterprise customers can request additional capabilities beyond standard plans.

<AccordionGroup>
  <Accordion title="Bring your own data sources" icon="database">
    Connect professional databases, licensed reference data, internal datasets, or organization-specific knowledge sources.
  </Accordion>

  <Accordion title="Custom compute and pipelines" icon="gears">
    Request custom pipeline configurations, larger compute environments, dedicated job templates, or specialized infrastructure.
  </Accordion>

  <Accordion title="Security and administration" icon="shield-check">
    Configure organization-level controls, support workflows, team administration, and enterprise deployment requirements.
  </Accordion>
</AccordionGroup>

<Tip>
  To discuss enterprise options, reach out to [contact@purna.ai](mailto:contact@purna.ai).
</Tip>