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

# AI Integrations

> Manage connections to external LLM providers like OpenAI, Azure, Bedrock, Vertex AI, and Anthropic within the Arize platform.

<Note>
  The `ai_integrations` client methods are currently in **ALPHA**. The API may change without notice. A one-time warning is emitted on first use.
</Note>

Manage connections to external LLM providers (OpenAI, Azure OpenAI, AWS Bedrock, Vertex AI, Anthropic, and others) for use within the Arize platform.

## Key Capabilities

* List AI integrations with optional filtering by space
* Retrieve integration details by name or ID
* Create integrations for any supported LLM provider
* Update integration settings without replacing the entire resource
* Delete integrations

## Supported Providers

| Provider value | Description      |
| -------------- | ---------------- |
| `openAI`       | OpenAI           |
| `azureOpenAI`  | Azure OpenAI     |
| `awsBedrock`   | AWS Bedrock      |
| `vertexAI`     | Google Vertex AI |
| `anthropic`    | Anthropic        |
| `gemini`       | Google Gemini    |
| `nvidiaNim`    | NVIDIA NIM       |
| `custom`       | Custom provider  |

## List AI Integrations

List all AI integrations you have access to, with optional filtering by space or name.

```python theme={null}
resp = client.ai_integrations.list(
    space="your-space-name-or-id",  # optional
    name="openai",                  # optional substring filter
    limit=50,
)

for integration in resp.ai_integrations:
    print(integration.id, integration.name, integration.provider)
```

For details on pagination, field introspection, and data conversion (to dict/JSON/DataFrame), see [Response Objects](/api-clients/python/version-8/overview#response-objects).

## Create an AI Integration

Integration names must be unique within the account.

### OpenAI

```python theme={null}
from arize.ai_integrations.types import AiIntegrationProvider

integration = client.ai_integrations.create(
    name="my-openai",
    provider=AiIntegrationProvider.OPENAI,
    api_key="sk-...",
    model_names=["gpt-4o", "gpt-4o-mini"],
)

print(integration.id)
```

### Azure OpenAI

```python theme={null}
from arize.ai_integrations.types import AiIntegrationProvider

integration = client.ai_integrations.create(
    name="my-azure-openai",
    provider=AiIntegrationProvider.AZUREOPENAI,
    api_key="your-azure-key",
    base_url="https://your-resource.openai.azure.com/",
    model_names=["gpt-4o"],
)
```

### AWS Bedrock

For AWS Bedrock, `provider_metadata` must include the `"aws"` `kind` discriminator and a `role_arn`.

```python theme={null}
from arize.ai_integrations.types import AiIntegrationProvider

integration = client.ai_integrations.create(
    name="my-bedrock",
    provider=AiIntegrationProvider.AWSBEDROCK,
    provider_metadata={
        "kind": "aws",
        "role_arn": "arn:aws:iam::123456789012:role/my-role",
    },
    model_names=["anthropic.claude-3-5-sonnet-20241022-v2:0"],
)
```

### Vertex AI

For Vertex AI, `provider_metadata` must include the `"gcp"` `kind` discriminator and `project_id`, `location`, and `project_access_label`.

```python theme={null}
from arize.ai_integrations.types import AiIntegrationProvider

integration = client.ai_integrations.create(
    name="my-vertex-ai",
    provider=AiIntegrationProvider.VERTEXAI,
    provider_metadata={
        "kind": "gcp",
        "project_id": "my-gcp-project",
        "location": "us-central1",
        "project_access_label": "my-label",
    },
    model_names=["gemini-2.0-flash"],
)
```

## Get an AI Integration

Retrieve a specific integration by name or ID.

```python theme={null}
integration = client.ai_integrations.get(integration="your-integration-name-or-id")

print(integration.id, integration.name)
```

## Update an AI Integration

Only the fields you pass are updated — omitted fields are left unchanged. To explicitly clear a nullable field (e.g. `api_key`), pass `None`.

```python theme={null}
integration = client.ai_integrations.update(
    integration="your-integration-name-or-id",
    name="updated-name",
    model_names=["gpt-4o", "gpt-4o-mini", "o3-mini"],
)

print(integration.name)
```

## Delete an AI Integration

Delete an integration by name or ID. This operation is irreversible.

```python theme={null}
client.ai_integrations.delete(integration="your-integration-name-or-id")

print("Integration deleted")
```