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

# Voyager Tokens

> Understanding token consumption in Voyager

## What are tokens

Tokens are how Cloud Voyager measures usage for select, compute-intensive operations. This guide explains what consumes a token, what does not, and how tokens work over the course of your contract.

<Note>
  Tokens do not apply to On-Prem scanners. On-Prem customers pay for their compute hardware directly instead. Because On-Prem compute capacity is fixed, it's still worth factoring your hardware's throughput into how you plan recipes and applications.
</Note>

### Token Expiration

Tokens expire on the annual anniversary of your contract. Unused tokens do not roll over to the next contract year.

## How tokens work

Tokens are not tied to every action you take in Voyager, only to a specific set of higher-value operations. Each of these operations has a **fixed token cost**, regardless of how long it takes to process.

For example, a standard reconstruction always costs 1 token, whether it completes in 5 minutes or 20. This gives you a predictable, consistent way to plan and budget your usage.

**Token consumption scales with the value of the analysis being performed.** Routine product actions remain available without token usage, while more compute-intensive analyses use tokens in proportion to the depth and value of the results they produce.

### Token consuming actions

<table>
  <colgroup>
    <col width="223" />

    <col width="98" />

    <col width="396" />
  </colgroup>

  <thead>
    <tr>
      <th>Action</th>
      <th>Tokens</th>
      <th>Notes</th>
    </tr>
  </thead>

  <tbody>
    <tr>
      <td>Standard Reconstruction</td>
      <td>1</td>
      <td>Includes the default reconstruction</td>
    </tr>

    <tr>
      <td>Helical Reconstruction</td>
      <td>5</td>
      <td>For high-resolution scans of tall/long aspect-ratio parts; opt-in on the scanner</td>
    </tr>

    <tr>
      <td>Standard Corrections</td>
      <td>1</td>
      <td>Includes AoRC, BHC, etc.</td>
    </tr>

    <tr>
      <td>Full-Resolution Mesh</td>
      <td>1</td>
      <td>Charged when a full-resolution mesh is generated</td>
    </tr>

    <tr>
      <td>Wall Thickness Analysis</td>
      <td>1</td>
      <td>Charged when the analysis is executed</td>
    </tr>

    <tr>
      <td>Full-Resolution Export</td>
      <td>1</td>
      <td>Charged when a full-resolution export is executed; supports *NRRD/API/CSM*</td>
    </tr>

    <tr>
      <td>Metal Artifact Reduction</td>
      <td>2</td>
      <td>In Voyager Labs</td>
    </tr>
  </tbody>
</table>

### Zero token actions

Actions that do not consume tokens include:

* Creating ROIs
* Creating Normal Meshes
* Creating Inspection Planes
* Creating a Full Resolution Plane
* Porosity Analysis
* Crack Analysis
* Inclusion Analysis
* Auto Dimensioning
* Renders
* Labs

<Note>
  If an action is not in the table above or on the zero token actions list, token usage is still being determined and no tokens will be charged for the time being.
</Note>

### Examples

* An auto-scan with a default standard reconstruction: **1 token**
* An auto-scan with a default standard reconstruction, followed by two additional manual reconstructions in the Voyager frontend: **3 tokens**
* A standard reconstruction followed by a full-resolution export: **2 tokens**
* A helical reconstruction: **5 tokens**
* Creating an ROI, a normal mesh, and an inspection plane: **0 tokens**

## Recipes

Recipes do not have a single fixed token price. A Recipe's overall token cost is dependent on its chargeable operations. To estimate the token cost of a recipe, add up the current cost of each chargeable operation it contains. Your Lumafield contact can help confirm an estimate for any recipe, especially one that includes a newer or less common operation.


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