Skip to content

feat: add Gemini 3 Pro to Google provider model list (#405) - #521

Open
mvanhorn wants to merge 5 commits into
AsyncFuncAI:mainfrom
mvanhorn:feat/405-deepwiki-405-gemini-3-pro
Open

mvanhorn wants to merge 5 commits into
AsyncFuncAI:mainfrom
mvanhorn:feat/405-deepwiki-405-gemini-3-pro

Conversation

@mvanhorn

Copy link
Copy Markdown

Adds gemini-3-pro to the Google provider model list per #405.

Changes:

  • api/config/generator.json - new gemini-3-pro entry in providers.google.models with the same temperature/top_p/top_k shape as the existing 2.5 entries
  • api/api.py:217-221 - mirror the addition in the exception-path fallback ModelConfig so 3 Pro still surfaces in the dropdown if load_generator_config() raises

default_model for the Google provider stays at gemini-2.5-flash, so existing users see no behavior change. Users opt into 3 Pro explicitly via the model picker.

Pattern matches PR #420 (Bedrock model additions) and the original 2.5 family additions in PR #318.

Fixes #405

Adds gemini-3-pro to providers.google.models in api/config/generator.json
with the same temperature/top_p/top_k shape as the existing 2.5 entries,
and mirrors the addition in the fallback ModelConfig in api/api.py:220
so the new model surfaces in the dropdown even if config load fails.

default_model remains gemini-2.5-flash so existing users see no behavior
change. Users opt into 3 Pro explicitly via the model picker.

Fixes AsyncFuncAI#405

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request adds support for the Gemini 3 Pro model by updating the model list in the API and providing its configuration in generator.json. Feedback was provided to reorder the models in the API fallback list to ensure consistency with the configuration file and prevent potential issues with default model selection in the frontend.

Comment thread api/api.py Outdated
@jiaxi-xu-fsx

Copy link
Copy Markdown

Small follow-up: I think gemini-3-pro is already superseded by gemini-3.1-pro-preview in the latest Gemini API docs. We should probably switch to gemini-3.1-pro-preview (or include both for compatibility).

Reference: https://ai.google.dev/gemini-api/docs/models/gemini-3.1-pro-preview

Per @jiaxi-xu-fsx's follow-up on AsyncFuncAI#405: gemini-3.1-pro-preview supersedes
gemini-3-pro in the latest Gemini API docs. Adding it alongside gemini-3-pro
so existing users on the older alias keep working while new users can opt
into the preview model.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@mvanhorn

Copy link
Copy Markdown
Author

Good catch, thanks! Pushed f9532d8 adding gemini-3.1-pro-preview to the google provider model list alongside gemini-3-pro. Kept both so users on the older alias keep working while anyone wanting the newer preview can opt in.

Left gemini-3-pro (not gemini-3.1-pro-preview) in the api.py error-fallback path since that's the "configs failed to load" branch and a stable model is the safer default there.

@mvanhorn

Copy link
Copy Markdown
Author

Merged main and slimmed this down (head 43edc72).

  • Main split api/api.py into routers, and its error fallback in api/routers/system.py now lists only gemini-2.5-flash. I left that as is, so the fallback-order review note no longer applies.
  • @jiaxi-xu-fsx was right that the 3 Pro entry should go. gemini-3-pro was never a Gemini API model id (the preview was gemini-3-pro-preview), and Google's models page now lists gemini-3-pro-preview as shut down. I removed it.

The PR now just adds gemini-3.1-pro-preview to the Google models in api/config/generator.json.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Add Gemini 3 Pro to LLM List

2 participants