-
on aistudio gemini-flash-latest was pointing to gemini-2.5-flash-XXX models previously, but it started to point to gemini-3-flash-preview recently(I realized today) which started to support thinking_level instead of thinking_budget. see #1 in sc:
-
see the mismatch in #2-3 in sc:
I believe it’s probably forgotten and I wanted to remind you for further conflicts.
Note: there is no problem on it’s working but there might be issues on assigning correct thinking settings.
