### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### What happened?

When using the OpenRouter Auto model (`openrouter/auto`) through LiteLLM, the response only returns `openrouter/auto` as the model name.

However, the actual model selected by OpenRouter's routing mechanism is not exposed in the LiteLLM response. When calling `openrouter/auto` directly through the OpenRouter API, the response includes the actual routed model name. This information appears to be unavailable when the same request is routed through LiteLLM.

This makes it difficult to:

* Track which model handled a request * Monitor routing decisions * Analyze model performance and costs * Debug auto-routing behavior

It would be helpful if LiteLLM could expose the actual model selected by OpenRouter in addition to the requested model (`openrouter/auto`).

For example:

```json { "model": "openrouter/auto", "routed_model": "anthropic/claude-sonnet-4" } ```

This would provide better visibility into OpenRouter's routing decisions and improve observability for applications using automatic model routing.

### Steps to Reproduce

1. Get openrout…