Adding Providers

ClawNex supports 15 provider types for connecting to LLM backends. Providers are managed in Configuration > Model Providers.

Supported Providers

LM Studio, OpenRouter, OpenAI, Anthropic, Azure OpenAI, Google Vertex, AWS Bedrock, Mistral, Groq, Together AI, Fireworks, Replicate, Ollama, HuggingFace, and Custom/Other.

Adding a Provider

1. Go to Configuration > Model Providers

2. Click Add Provider

3. Select LM Studio as the type

4. Enter the API base URL (e.g., http://192.168.1.100:1234/v1)

5. API key: use lmstudio-local (static)

6. Configure model settings (context window, capabilities)

7. Click Save

8. Go to Configuration > Model Providers

9. Click Add Provider

10. Select OpenAI or Anthropic

11. Enter your API key from the provider dashboard

12. The API base URL is pre-filled for cloud providers

13. Click Save

14. Go to Configuration > Model Providers

15. Click Add Provider

16. Select OpenRouter

17. Enter your OpenRouter API key from openrouter.ai

18. Click Save

OpenRouter gives you access to 100+ models through a single API key.

LiteLLM Configuration

Providers configured through the dashboard are synced to litellm/config.yaml automatically. For manual configuration, edit:

• ~/sentinel/litellm/config.yaml — update api_base URLs and model list
• ~/sentinel/litellm/start.sh — update API key exports

Testing the Connection

After adding a provider:

1. Go to Infrastructure panel
2. Verify the LiteLLM Proxy shows ONLINE (green)
3. Go to Prompt Shield and run a manual scan — if it returns results, the shield engine is working
4. Send a test message through an agent to confirm traffic flows through the proxy