PR #21642: Google LLMs: Error earlier if file is too large with vertexai client

• Repository: run-llama/llama_index
• Author: connorszczepaniak-wk
• State: open | merged: False
• Link: https://github.com/run-llama/llama_index/pull/21642

Description (problem / solution / changelog)

Description

When interacting with Google LLMs and using vertexai, uploading a file >20MB results in a cryptic error. Since we know the google client will fail when we fall through, let's proactively error instead.

Fixes #21639

New Package?

Did I fill in the tool.llamahub section in the pyproject.toml and provide a detailed README.md for my new integration or package?

• Yes
• No

Version Bump?

Did I bump the version in the pyproject.toml file of the package I am updating? (Except for the llama-index-core package)

• Yes
• No

Type of Change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Your pull-request will likely not be merged unless it is covered by some form of impactful unit testing.

• I added new unit tests to cover this change
• I believe this change is already covered by existing unit tests

Suggested Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added Google Colab support for the newly added notebooks.
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I ran uv run make format; uv run make lint to appease the lint gods

Changed files

• llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/utils.py (modified, +4/-0)
• llama-index-integrations/llms/llama-index-llms-google-genai/tests/test_llms_google_genai.py (modified, +14/-0)

PR #21644: fix(google-genai): error early on Vertex AI File API uploads

• Repository: run-llama/llama_index
• Author: Tajaddin
• State: open | merged: False
• Link: https://github.com/run-llama/llama_index/pull/21644

Description (problem / solution / changelog)

Description

The Vertex AI Gemini API does not support file uploads through the Gemini File API. In create_file_part (llama-index-llms-google-genai), any code path calling client.aio.files.upload() fails on a Vertex client with the cryptic error from google-genai:

This method is only supported in the Gemini Developer client.

The two cases hitting this on Vertex are:

• file_mode='fileapi' for files of any size
• file_mode='hybrid' with files larger than 20MB

Both used to surface the confusing message deep inside the upload call. Users get no signal indicating the real cause (Vertex does not expose this API at all) instead of a misconfiguration on their side.

This PR adds a pre-flight check on client.vertexai and raises a clear ValueError with a remediation path before the upload is attempted. Inline mode under 20MB continues to work on Vertex unchanged.

Fixes #21639

New Package?

• Yes
• No

Version Bump?

Bumped llama-index-llms-google-genai from 0.9.3 to 0.9.4.

• Yes
• No

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Added three unit tests under tests/test_llms_google_genai.py:

• test_create_file_part_raises_on_vertexai_with_fileapi — fileapi mode + Vertex client → ValueError, no upload attempted.
• test_create_file_part_raises_on_vertexai_with_large_hybrid_file — 21MB hybrid + Vertex → ValueError, no upload attempted.
• test_create_file_part_inline_still_works_on_vertexai — inline mode + Vertex + small file → returns Part, no upload attempted (regression guard).

Also updated the two pre-existing create_file_part tests to set mock_client.vertexai = False, since a bare MagicMock auto-generates a truthy vertexai attribute.

Local results:

• uv run -- pytest tests/ → 33 passed, 46 skipped (skips need GOOGLE_API_KEY / GOOGLE_GENAI_USE_VERTEXAI env vars).

• uv run -- ruff check on the changed files → clean.

• uv run -- ruff format --check on the changed files → clean.

• codespell on the changed files → clean.

• I added new unit tests to cover this change

• I believe this change is already covered by existing unit tests

Suggested Checklist

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation (CHANGELOG entry)
• I have added Google Colab support for the newly added notebooks (N/A, no notebooks)
• My changes generate no new warnings
• I have added tests that prove my fix is effective
• New and existing unit tests pass locally with my changes
• I ran ruff check and ruff format on the modified files

AI assistance disclosure (per CONTRIBUTING.md)

Drafted with Claude Code assistance. The fix is a 10-line guard plus three unit tests. I reviewed every line, traced the google-genai failure mode against the SDK source linked in the issue, and ran the full local test suite and the linters.

Changed files

• llama-index-integrations/llms/llama-index-llms-google-genai/CHANGELOG.md (modified, +6/-0)
• llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/utils.py (modified, +10/-0)
• llama-index-integrations/llms/llama-index-llms-google-genai/pyproject.toml (modified, +1/-1)
• llama-index-integrations/llms/llama-index-llms-google-genai/tests/test_llms_google_genai.py (modified, +75/-0)

PR #21646: fix: reject Vertex AI file uploads in google genai

• Repository: run-llama/llama_index
• Author: he-yufeng
• State: open | merged: False
• Link: https://github.com/run-llama/llama_index/pull/21646

Description (problem / solution / changelog)

## Summary

• raise a clear ValueError before Google GenAI File API uploads are attempted with a Vertex AI client
• keep small hybrid files on the inline path
• add regression coverage for fileapi mode and large hybrid files with Vertex AI

Fixes #21639.

To verify

• python -m py_compile llama_index\llms\google_genai\utils.py tests\test_llms_google_genai.py
• python -m ruff check llama_index\llms\google_genai\utils.py tests\test_llms_google_genai.py
• python -m ruff format --check llama_index\llms\google_genai\utils.py tests\test_llms_google_genai.py
• python -m pytest tests\test_llms_google_genai.py -q

Changed files

• llama-index-integrations/llms/llama-index-llms-google-genai/llama_index/llms/google_genai/utils.py (modified, +12/-5)
• llama-index-integrations/llms/llama-index-llms-google-genai/tests/test_llms_google_genai.py (modified, +30/-0)