Update Gemini API URL and model version#5454
Conversation
Summary of ChangesHello, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request addresses breaking changes in the Gemini API integration by reverting to the v1beta endpoint and updating the default model version. These changes ensure that the gap analyzer tool remains functional with the latest Google AI service requirements. Highlights
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize the Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counterproductive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here. Footnotes
|
There was a problem hiding this comment.
Code Review
This pull request updates the Gemini API URL to the v1beta version and changes the default model identifier to gemini-3.1-pro-preview. Feedback indicates that the specified model name is invalid and will result in API failures, suggesting a revert to a stable identifier like gemini-1.5-pro.
Corrected a typo in the README regarding telemetry paths.
Seems like 2.5-pro has been deprecated on the v1beta endpoint which caused the first error.
The v1 endpoint's GenerationConfig object currently lacks the responseMimeType and responseSchema fields entirely, causing the second error. This is why v1beta was used to begin with.
As with most AI tools, the model/name in the code will have to be updated every time a major release is made by Google. Modified the model to point to 3.1-pro-preview and reverted the v1 change to v1beta.