Why can’t I attach a PDF reference material (such as a scientific paper or a part of a handbook) to a dialog to help the model write the code without me carrying it over to Google Colab for running every time? The code execution toggle-switch grays out immediately once I do this and the model isn’t able to run its own code anymore, even if it was in the previous turns of the conversation
Thanks for bringing this up , we are abel to see this on our end aswell,
we will update the gemini team on this,
Thankyou
Five months later, the problem still persists. Do you thunk you could provide any update?
Even after a year and a month, the problem is still there. And it is not even limited to AI Studio anymore. It is also present in the Gemini app, where it makes the agent practically useless and stupid.
And Google still has not fixed it. Or I should say, they apparently do not know how to fix it, considering they have known about the problem for over a year and still seem to be ignoring it.
Imagine finding out that they already knew about this problem and even acknowledged it publicly. I only found out because of ChatGPT, since I barely use Gemini anymore because of how incompetent the people working on it seem to be.
When are you guys actually planning to fix this? Or does user feedback just not matter to you at all?
I guess this is not an AI Studio problem but a Gemini API problem, When they enabled Google Search and Google Maps simultaneously (even though only for Flash models, those two are still incompatible for 3.1 Pro) in API a month ago it became available in AI Studio as well. Also note: “Been on the backlog for a long time”…
It seems vibecoding takes a priority compared to API and chats, but then if you use Antigravity you must pay for Google Cloud Document AI