# Define AI Provider

Nected AI Copilot uses Claude by default and includes up to 5 free chats. To use Copilot beyond the free limit — or to use your own Claude model — you need to connect an AI provider integration.

There are two ways to add an AI provider: through the **Integrations page** or directly from the **Copilot chat window**.

### Method 1: Add AI Provider from the Integrations Page

This is the recommended path for setting up a new provider from scratch.

**Step 1:** From the left navigation, go to **Integrations**.

**Step 2:** On the Integrations page, click **Add New Integration** or the **+** button.

<figure><img src="/files/KF8GeKk5EeCw9FBw2wLG" alt=""><figcaption></figcaption></figure>

**Step 3:** In the integration type selector, choose **AI Model** (or search for Claude, OpenAI, or the model you want to connect).

\[Screenshot: Integration type selector with AI Model option]

**Step 4:** Fill in the required fields:

* **Integration name** — a label for this integration (for example, "My Claude API")
* **API Key** — your API key from the provider (for example, from Anthropic's console for Claude)
* **Model** — select the specific model version you want to use

<figure><img src="/files/QgdirbJArR5zJVIR47kw" alt=""><figcaption></figcaption></figure>

**Step 5:** Click **Save**. The integration will be created in **Draft** state by default.

**Step 6:** To make this integration available in the Copilot chat window, you must **Publish** it. Click the **Publish** button on the integration card or detail view.

{% hint style="success" icon="sparkles" %}
**Important:** Only published integrations appear in the Copilot chat window. Draft integrations are not shown.
{% endhint %}

Once published, the integration is ready to be selected inside the Copilot chat window.

### Method 2: Add or Select AI Provider from the Copilot Chat Window

If you already have a published integration and want to activate it for Copilot, or if you want to quickly jump to creating a new one, you can do this directly from the chat.

**Step 1:** Open **Nected AI Copilot Beta** from the **Home** page.

**Step 2:** In the chat window, click the **Settings** (gear) icon. This opens the Copilot settings panel.

<figure><img src="/files/vp0mf8AvFT9mpQ2zjyWb" alt=""><figcaption></figcaption></figure>

**Step 3:** In the settings panel, you can:

* Set the **Max Tokens** limit for responses
* Select an **Integration** from the dropdown, this list shows all your **published** AI model integrations

**Step 4:** Select the integration you want to use from the dropdown and confirm your settings.

{% hint style="info" %}
**Note:** Only published integrations are listed here. If you do not see your integration in the dropdown, go back to the Integrations page and make sure it has been published.
{% endhint %}

**Step 5 (Optional): Create a new integration from here.** If you want to add a new provider without leaving the chat, click the **+** (plus) icon next to the integration dropdown. This takes you directly to the Integrations page where you can create and publish a new model integration. Once published, return to the chat settings to select it.

### After Setup

Once an integration is selected, Copilot will use that model for all subsequent chats in your workspace. You can change the integration at any time by returning to the chat settings panel.

> **Recommendation:** Nected's Rules and Workflows are optimized for Claude. If you are connecting your own model, use a Claude-based model for the best results.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.nected.ai/nected-docs/ai-assistant-copilot/define-ai-provider.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.