Model Deprecation – [API Catalog] [deepseek-v4-pro] [Missing Notice]

## Model Deprecation – [API Catalog] [deepseek-v4-pro] [Missing Notice]

### Summary
The model `deepseek-ai/deepseek-v4-pro` is still listed in the API catalog but returns HTTP 410 Gone.

### Current Behavior
- **Model listed:** `deepseek-ai/deepseek-v4-pro` appears in `GET /v1/models`
- **Actual response:** `410 Gone` with message:
```json
{
"type": "about:blank",
"title": "Gone",
"status": 410,
"detail": "The model 'deepseek-ai/deepseek-v4-pro' has reached its end of life on 2026-08-07T09:00:00Z and is no longer available."
}
```

### Expected Behavior
The model should be removed from the catalog listing, or marked as deprecated/unavailable in the metadata.

### Affected Model
- **Deprecated model:** `deepseek-ai/deepseek-v4-pro`
- **Replacement:** `deepseek-ai/deepseek-v4-pro-07313` (works correctly)
- **EOL Date:** 2026-08-07

### How to Reproduce
```bash
curl -s https://integrate.api.nvidia.com/v1/models | jq '.data[] | select(.id | endswith("deepseek-v4-pro"))' | .id
```

### Additional Context
This affects downstream tools like OpenCode that parse the model catalog. The deprecated model creates false positives in model selection.

Start by reproducing the catalog response with the provided curl and jq command, then locate where the API catalog sources or exposes model metadata. Check how `deepseek-ai/deepseek-v4-pro` is represented and compare it with `deepseek-ai/deepseek-v4-pro-07313`. Done means the deprecated model is removed from `GET /v1/models` or clearly marked unavailable, with coverage for the catalog behavior if tests exist.

Written by the indexing model from the issue text.