### Check for existing issues

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

### What happened?

Streaming requests to Vertex AI Anthropic models (e.g. `claude-sonnet-4-5@20250929`) via the `/v1/messages` endpoint fail with 404 when a custom `api_base` is configured in `litellm_params`. The required `:streamRawPredict` suffix is not appended to the URL.

The proxy logs an error ``` HTTPStatusError("Client error '404 Not Found' for url 'https://aiplatform.us.rep.googleapis.com/v1/projects/my-project/locations/us/publishers/anthropic/models/claude-sonnet-4-5@20250929'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404") ```

## Root cause

Commit `cdb4917185` (PR #21658) introduced a regression in `litellm/llms/vertex_ai/vertex_ai_partner_models/anthropic/experimental_pass_through/transformation.py`.

The fix moved the `get_complete_vertex_url()` call behind an `if api_base is None:` guard. This means when a user provides a custom `api_base` in their litellm config, `get_complete_vertex_url()` is never called and the required endpoint suffix (`:streamRawPredict`) is never appended.

Before the commit: `…