Fixing test_parse_office_doc by modernizing Gemini model

[jamesbraza]

From https://developers.googleblog.com/gemini-embedding-available-gemini-api/:

Legacy models will also be deprecated in the coming months:

• ...
• text-embedding-004 on January 14, 2026

This PR just modernizes the model for test_parse_office_doc to the latest gemini/gemini-embedding-001

---

Note

Low Risk
Test-only change that updates a model identifier; main risk is CI flakiness if the new embedding model behaves differently or has different availability/quotas.

Overview
Updates test_parse_office_doc to use gemini/gemini-embedding-001 instead of the deprecated gemini/text-embedding-004 embedding model, keeping the rest of the Gemini LLM configuration unchanged.

^{Written by Cursor Bugbot for commit b8c7969. Configure here.}

[dosubot]

Related Documentation

Checked 1 published document(s) in 1 knowledge base(s). No updates required.

^{How did I do? Any feedback?}  

[Copilot]

Pull request overview

Updates the Gemini embedding model used by test_parse_office_doc to address a failing test by modernizing the model configuration.

Changes:

• Switch test_parse_office_doc embedding model from gemini/text-embedding-004 to gemini/gemini-embedding-001.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The PR description references Gemini embedding model text-embedding-004, but this change switches the test to gemini-embedding-001. This is inconsistent with the rationale given in the PR description and also diverges from the README’s Gemini example (README.md:578 uses gemini/text-embedding-004). Please either (a) update the test to the intended model from the PR description, or (b) update the PR description (and ideally the README example in a follow-up/this PR) to match the new recommended embedding model name.