feat(google-vertex/google/gemini-3-pro-image): add new models [bot]#1192
Conversation
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
|
/test-models |
Gateway test results
Skipped (1)
Skip reason |
…age-20260530-000623
|
/test-models |
harshiv-26
deleted the
bot/add-google-vertex-google-gemini-3-pro-image-20260530-000623
branch
Gateway test results
Skipped (1)
Skip reason |
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 2 potential issues.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 2649773. Configure here.
| - function_calling | ||
| - system_messages | ||
| - structured_output | ||
| - tool_choice |
There was a problem hiding this comment.
False tool support advertised
Medium Severity
The config lists function_calling and tool_choice under features, but Gemini 3 Pro Image on Vertex does not enable function calling, and every other google/gemini-*-image entry here only declares system_messages and uses removeParams to drop tool_choice from the provider default.
Reviewed by Cursor Bugbot for commit 2649773. Configure here.
| modalities: | ||
| input: | ||
| - text | ||
| - image |
There was a problem hiding this comment.
Missing PDF input modality
Medium Severity
Input modalities only list text and image, but Vertex documents PDF as a supported input for Gemini 3 Pro Image, and the adjacent google/gemini-3-pro-image-preview entry on the same provider includes pdf.
Reviewed by Cursor Bugbot for commit 2649773. Configure here.
1 participant
Auto-generated by model-addition-agent for
google-vertex/google/gemini-3-pro-image.Note
Low Risk
Metadata-only addition with no application logic changes; incorrect pricing or limits would affect billing/routing display only.
Overview
Adds a new Google Vertex model definition at
providers/google-vertex/google/gemini-3-pro-image.yamlforgoogle/gemini-3-pro-image, enabling routing, pricing, and capability metadata for Gemini 3 Pro Image on thegoogle/provider path.The entry marks the model as preview with image primary mode and chat in
supportedModes, thinking enabled, and multimodal text/image input and output. It includes global token and per-image costs, a 65536 context window, and features such as function calling, structured output, and tool choice (unlike the siblinggoogle/gemini-3-pro-image-previewconfig, which removestool_choice).Reviewed by Cursor Bugbot for commit 2649773. Bugbot is set up for automated code reviews on this repo. Configure here.