### Check for existing issues

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

### What happened?

When configuring custom model pricing via `model_info` in the LiteLLM config YAML, the `*_above_200k_tokens` tiered pricing fields are silently ignored during cost calculation. The base rate is applied flat across all tokens regardless of request size, resulting in significant cost undercharging for large-context requests

### Steps to Reproduce

1. Configure a model with above-200k tiered pricing in `model_info`:

```yaml - model_name: "claude-opus-4-6" litellm_params: model: vertex_ai/claude-opus-4-6 vertex_project: "my-project" vertex_location: "global" model_info: input_cost_per_token: 0.000005 output_cost_per_token: 0.000025 cache_read_input_token_cost: 0.0000005 cache_creation_input_token_cost: 0.00000625 input_cost_per_token_above_200k_tokens: 0.00001 output_cost_per_token_above_200k_tokens: 0.0000375 cache_creation_input_token_cost_above_200k_tokens: 0.0000125 cache_read_input_token_cost_above_200k_tokens: 0.000001

Send a request with >200k prompt tokens (e.g. 585,659 cache creation token…