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

# TypeSafe AI

> Add a TypeSafe AI API key to run Decision Judge evaluations on the Jev decision model in Arize AX

Add your TypeSafe AI API key so Arize AX can call **Jev** when you run [Decision Judge](/docs/ax/evaluate/decision-judge) evaluators.

TypeSafe is a **restricted-use, eval-only** provider: it powers Decision Judge evaluators and does not appear in the prompt playground, LLM-as-a-judge model pickers, managed-agent model pickers, or the `/v2/ai-integrations` REST API. There is no base URL or model list to configure. Arize always calls `jev-latest`.

> [!NOTE]
>
> By adding this integration, your data may be sent to TypeSafe AI when your Decision Judge evaluators run, and your TypeSafe AI account will be billed for usage.

## Add the Integration from Arize AX

**Navigate to Settings → AI Providers**, then select **TypeSafe AI**.

![TypeSafe AI integration slideover with name, API key, successful Test Integration result, and terms checkbox](https://storage.googleapis.com/arize-assets/doc-images/typesafe_ai_provider.png)

1. Give the integration a **name** and paste your TypeSafe **API key**.
2. Accept the terms checkbox.
3. Optionally click **Test Integration** (billable). A successful test shows **TypeSafe AI responded successfully.**
4. Save the integration.

Creating AI Providers requires Organization Admin or Account Admin access. After saving, create a [Decision Judge](/docs/ax/evaluate/decision-judge) evaluator and select this integration.