# Configuring and Enabling AI Providers in SecureLLM

This tutorial shows how administrators can configure and enable AI providers in SecureLLM to make models available through the gateway.

Providers define which external or internal model services SecureLLM can route requests to. Administrators can add providers, control allowed models, and enable or disable providers as needed.

---

## Tutorial Overview

In this tutorial, an administrator:

- Adds a new AI provider  
- Configures provider credentials and endpoint settings  
- Defines allowed models  
- Enables the provider for use  
- Verifies models are available to users  

---

## Prerequisites

Before you begin, ensure:

- You have administrator access in SecureLLM  
- You have provider connection details available  
- You have an API key or authentication credential for the provider  
- You know which models should be exposed to users  

Examples of provider types:

- Commercial model providers  
- Internal model endpoints  
- Private model services  

---

## Tutorial Steps

### Step 1: Open Providers

1. Log into SecureLLM.
2. From the sidebar, select **Providers**.
3. Review configured providers or add a new one.

---

### Step 2: Add a Provider

1. Click **Add Provider**.
2. Select the provider type.
3. Enter required configuration details:

- Provider API key  
- Base URL (if applicable)  
- Provider name  
- Authentication settings  

Example:

```text
Provider Type: OpenAI-Compatible Endpoint
Base URL: https://api.example.com/v1
```

4. Click **Save**.

---

### Step 3: Configure Allowed Models

Optionally restrict which models are exposed through this provider.

Example:

```text
gpt-4
claude-sonnet
embedding-model-v1
```

You can:

- Allow all models  
- Restrict to specific models  
- Limit access to approved models only  

Save configuration changes.

---

### Step 4: Enable the Provider

1. Locate the provider in the Providers page.
2. Use the toggle switch to enable it.
3. Confirm the provider is active.

Once enabled, SecureLLM can route requests to this provider.

---

### Step 5: Verify Model Availability

1. Open **Models**.
2. Search or filter by the provider.
3. Verify configured models appear in the list.

Confirm users can discover models from the provider.

---

### Step 6: Test Requests Through the Provider

Submit a test request using one of the provider’s models.

Verify:

- Request succeeds  
- Response is returned  
- Usage is recorded  
- No routing or authentication errors occur  

---

## Managing Provider Availability

Providers can be enabled or disabled without deleting configuration.

### Disable a Provider

Use the provider toggle to temporarily stop routing requests.

Common reasons:

- Maintenance  
- Cost control  
- Provider instability  
- Policy changes  

---

### Delete a Provider

Delete a provider only if it will no longer be used.

Review dependencies before removal.

---

## Expected Outcome

After completing this workflow, you should be able to:

- Add and configure AI providers in SecureLLM  
- Control which models are available  
- Enable providers for request routing  
- Verify model access for users  
- Manage provider availability over time  

---

## Troubleshooting

### Provider Fails Authentication

Check:

- API key is valid  
- Credentials were entered correctly  
- Base URL is correct  
- Provider endpoint is reachable  

---

### Models Do Not Appear in Models Page

Verify:

- Provider is enabled  
- Allowed models are configured correctly  
- Model list was refreshed  
- User access restrictions are not blocking visibility  

---

### Requests Fail After Provider Is Enabled

Review:

- Provider configuration  
- Routing configuration  
- Request logs  
- Provider-side availability  

---

## Related Tutorials

- Discovering and Selecting Models in SecureLLM  
- Managing User Access Controls in SecureLLM  
- Testing and Deploying Guardrail Policies in SecureLLM  
- Monitoring Gateway Health and Performance in SecureLLM