### Check for existing issues

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

### The Feature

Support an optional deployment/model-level `cost_discount` field in custom pricing metadata.

Example proxy/router config:

```yaml model_list: - model_name: gpt-4o-mini-discounted litellm_params: model: openai/gpt-4o-mini api_key: os.environ/OPENAI_API_KEY model_info: id: discounted-deployment-1 cost_discount: 0.8 ```

`cost_discount` would be a float between `0` and `1`, where `0.8` means an 80% discount is applied to the calculated base cost. The final reported cost would be `base_cost * (1 - cost_discount)`.

### Motivation, pitch

LiteLLM already supports provider-level discounts through `cost_discount_config`, but some deployments need accurate spend tracking for negotiated or deployment-specific pricing without changing the shared base model price.

Today, operators can override input/output token prices per deployment, but there is no generic way to say "use the normal built-in model pricing, then apply this deployment's discount." This makes it harder to model negotiated pricing, reseller pricing, and per…