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

# Error Handling

> Learn about common errors and how to troubleshoot them

## What is Error Handling?

This guide helps you understand common errors you might encounter when using Model Router and how to resolve them.

## Common Errors and Solutions

### Authentication Errors

#### Error: "Invalid API Key" or 401 Unauthorized

**What it means**: Your API key is incorrect, missing, or invalid.

**How to fix**:

1. Check that your API key starts with `sk-`
2. Verify you copied the entire API key correctly
3. Make sure you're including it in the `Authorization` header: `Bearer sk-your-key`
4. If you lost your key, create a new one in the [console](https://app.memorylake.ai/panel/token)

#### Error: "API key not found"

**What it means**: The API key doesn't exist or has been deleted.

**How to fix**:

1. Verify the API key exists in the [console](https://app.memorylake.ai/panel/token)
2. Create a new API key if needed
3. Make sure you're using the correct API key

### Model Errors

#### Error: "Model not found" or "Model does not exist"

**What it means**: The model you're trying to use isn't available to your API key.

**How to fix**:

1. Check [List Available Models](/features/model-router/list-available-models) to see which models you can use
2. Verify the model name is spelled correctly (model names are case-sensitive)
3. Use the exact model `id` from the model list
4. Contact your administrator if you need access to a specific model

#### Error: "Model is not available"

**What it means**: The model exists but isn't enabled for your API key group.

**How to fix**:

1. Check which models are available to your API key
2. Contact your administrator to enable the model for your group
3. See [Limits and Prerequisites](/features/model-router/others/limits-and-prerequisites) for more information

### Quota Errors

#### Error: "Insufficient quota" or "Quota exceeded"

**What it means**: You don't have enough quota to make the request.

**How to fix**:

1. Check your balance in the console under **Billing → Overview**
2. [View your usage](/features/model-router/view-usage-and-billing) to see how much you've consumed
3. Top up credits, or ask your team admin to raise your member quota
4. Confirm you're in the intended space — personal and team contexts draw on different quota pools

<Info>
  Quota is checked before each request. If you don't have enough quota, the request will be rejected immediately. See [View Usage and Billing](/features/model-router/view-usage-and-billing) for details.
</Info>

### Rate Limiting Errors

#### Error: 429 "Too Many Requests" or "Rate limit exceeded"

**What it means**: You're making requests too quickly.

**How to fix**:

1. Wait a moment and try again
2. Reduce the frequency of your requests
3. Implement exponential backoff in your code
4. Batch requests if possible
5. See [Reliability and Failover](/features/model-router/others/reliability-and-failover) for retry strategies

### Request Format Errors

#### Error: "Invalid request format" or 400 Bad Request

**What it means**: The request body or parameters are incorrect.

**How to fix**:

1. Check that your request body is valid JSON
2. Verify all required fields are present
3. Make sure field names and types match the API specification
4. Review the [Direct API Requests](/features/model-router/getting-started/use-api-key/direct-api-requests) guide for correct format

#### Error: "Missing required field"

**What it means**: A required parameter is missing from your request.

**How to fix**:

1. Check the API documentation for required fields
2. Verify all required parameters are included
3. Check that field names are spelled correctly

### Network Errors

#### Error: "Connection timeout" or "Network error"

**What it means**: The request couldn't reach the server or took too long.

**How to fix**:

1. Check your internet connection
2. Verify the API endpoint URL is correct: `https://app.memorylake.ai`
3. Try again after a moment
4. Check if there are any network issues
5. See [Reliability and Failover](/features/model-router/others/reliability-and-failover) for automatic retry mechanisms

### Streaming Errors

#### Error: "Stream connection closed unexpectedly"

**What it means**: The streaming connection was interrupted.

**How to fix**:

1. This is normal for long streams - implement reconnection logic
2. Check your network stability
3. Handle stream interruptions gracefully in your code
4. See [Reliability and Failover](/features/model-router/others/reliability-and-failover) for streaming best practices

## Error Response Format

When an error occurs, the API returns a JSON response with error details:

```json theme={null}
{
  "error": {
    "message": "Model not found",
    "type": "invalid_request_error",
    "code": "model_not_found"
  }
}
```

## Best Practices for Error Handling

1. **Always check the response status**: Handle both success and error cases
2. **Read error messages**: The error message usually tells you what went wrong
3. **Implement retry logic**: For transient errors like rate limits or network issues
4. **Log errors**: Keep track of errors for debugging
5. **Handle gracefully**: Don't crash your application on API errors

## Example: Error Handling in Code

### Python Example

```python theme={null}
import requests

def make_request(api_key, model, messages):
    try:
        response = requests.post(
            "https://app.memorylake.ai/v1/chat/completions",
            headers={
                "Authorization": f"Bearer {api_key}",
                "Content-Type": "application/json"
            },
            json={
                "model": model,
                "messages": messages
            }
        )
        response.raise_for_status()
        return response.json()
    except requests.exceptions.HTTPError as e:
        if response.status_code == 401:
            print("Error: Invalid API key")
        elif response.status_code == 404:
            print("Error: Model not found")
        elif response.status_code == 429:
            print("Error: Rate limit exceeded. Please wait and try again.")
        else:
            print(f"Error: {response.status_code} - {response.text}")
        raise
    except requests.exceptions.RequestException as e:
        print(f"Network error: {e}")
        raise
```

## Getting Help

If you're still having issues:

1. Check the [Limits and Prerequisites](/features/model-router/others/limits-and-prerequisites) guide
2. Review the [Reliability and Failover](/features/model-router/others/reliability-and-failover) documentation
3. Check your API key and model availability
4. Contact [support@memorylake.ai](mailto:support@memorylake.ai) with:
   * The exact error message
   * Your API key (masked)
   * The request you were making
   * Any relevant logs

## Related Documentation

* [Limits and Prerequisites](/features/model-router/others/limits-and-prerequisites)
* [Reliability and Failover](/features/model-router/others/reliability-and-failover)
* [List Available Models](/features/model-router/list-available-models)
* [View Usage and Billing](/features/model-router/view-usage-and-billing)
* [Direct API Requests](/features/model-router/getting-started/use-api-key/direct-api-requests)