> For the complete documentation index, see [llms.txt](https://cogna8-io.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cogna8-io.gitbook.io/docs/settings/model-testing.md).

# Model Testing

Model Testing lets you compare how different LLMs perform at the state extraction stage of the integrity pipeline. This is useful for evaluating which models produce the most reliable state formation for your specific use case.

***

## How It Works

{% stepper %}
{% step %}
**Select Models**

Choose which LLMs to compare from the available models.
{% endstep %}

{% step %}
**Provide Input**

Enter a test conversation or select a pre-built test scenario.
{% endstep %}

{% step %}
**Compare Results**

See side-by-side extraction results: what each model identified, how it classified items, and where they differ.
{% endstep %}
{% endstepper %}

{% hint style="info" %}
**Model testing does not affect production conversations.** All testing runs in an isolated environment.
{% endhint %}