# AI Connections

## AI Provider Connections

### Description

An **AI Connection** (shown in the UI as an **AI Provider Connection**) is a Connection record that includes an `aiProvider` and the authentication details required to call that provider through Contextual’s AI Gateway. AI Connections are used by:

* [AI Routes](/documentation-and-resources/components-and-data/ai-routes.md)
* [AI Generate](/documentation-and-resources/components-and-data/flows/node-reference/ai-gateway/ai-generate.md)

### Creating an AI Connection

1. In your tenant workspace, go to **Components → Connections**
2. Click **Create Connection**
3. Choose **AI Provider Connection**
4. Provide a Connection ID, Connection Name, choose an AI Provider, and complete the provider-specific configuration
5. Save

### Required Information

All AI Connections require:

* **Connection ID** (`apiId`) - Lower case letters, numbers, and dashes (`^[a-z0-9-]+$`)
* **Connection Name** (`name`)
* **AI Provider** (`aiProvider`)

Provider-specific requirements:

#### OpenAI

* **API Key** (`bearerToken`)
* **Endpoint** (default: `https://api.openai.com/v1`)
* Underlying Connection type: `bearer`

#### Anthropic

* **API Key** (stored in headers as `x-api-key`)
* **Endpoint** (default: `https://api.anthropic.com/v1`)
* Underlying Connection type: `public`

#### Azure OpenAI

* **Resource Name** (`aiProviderData.resourceName`)
* **API Key** (stored in headers as `api-key`)
* **Endpoint** is generated from the Resource Name (default: `https://{resourceName}.openai.azure.com/openai`) and can be overridden
* Underlying Connection type: `public`

#### Google AI (Gemini)

* **API Key** (stored in headers as `x-goog-api-key`)
* **Endpoint** (default: `https://generativelanguage.googleapis.com/v1beta`)
* Underlying Connection type: `public`

#### Vertex AI

* **Project ID** (`aiProviderData.projectId`)
* **Location** (`aiProviderData.location`, for example `us-central1` or `global`)
* **Service Account Key (JSON)** (`key`)
* **Endpoint** is generated from Project ID + Location and can be overridden
* Underlying Connection type: `google-service-account` (uses scope `https://www.googleapis.com/auth/cloud-platform`)

#### Vertex AI Anthropic

Same requirements as Vertex AI, but routes to Anthropic models through Vertex AI’s Anthropic publisher endpoint. Anthropic models need to be explicitly enabled for each Vertex AI account, so make sure you are opted in to the particular Anthropic model you want to access through Vertex.

### Headers Note

Additional headers you add to a Connection are included in requests made by HTTP Nodes, but are **not** used in requests made through AI Routes / the AI Gateway.

### Usage

* AI Routes only allow selecting Connections that have an `aiProvider` configured (AI Connections).
* Use the Connection’s Endpoint as the base URL when making manual provider calls from HTTP nodes (if needed).
* AI Connections are standard Contextual Connections, so they can be used anywhere a Connection is accepted (for example, selecting the AI Connection in an HTTP Request/HTTP node to call the provider API directly without using an AI Route or the AI Generate node).


---

# 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.contextual.io/documentation-and-resources/components-and-data/connections/types-of-connections/ai-connections.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.