Description of the issue: The “Create dataset” and “Add to existing dataset” functionalities in the AI Studio “Logs and Datasets” tab are completely broken. When trying to create a new dataset from selected unassigned logs, the UI dialog freezes after clicking “Create a dataset”. There is no error message in the UI, but the browser’s DevTools console reveals that the frontend is trying to reach a dead endpoint.
This completely blocks essential workflows (like exporting prompt logs for time tracking and billing). I have spent a lot of time building a hardened workflow around this, and it is currently entirely unusable.
Technical Details (from Chrome DevTools Console): When clicking the button, the following errors are thrown immediately:
-
POST https://alkalimakersuite-pa.clients6.google.com/$rpc/google.internal.alkali.applications.makersuite.v1.MakerSuiteService/CreateDataset 404 (Not Found) -
Uncaught (in promise) RpcError: Requested entity was not found.
Steps to reproduce:
-
Go to “Logs and Datasets” in Google AI Studio.
-
Check the boxes to select unassigned logs.
-
Click the “Create dataset” button.
-
Enter a valid name (e.g., “TIMESHEET_04_2026”) and click “Create a dataset”.
-
Result: The popup stays open, nothing happens in the UI, and the 404 error is logged in the console.
Impact: As a paying user, not being able to access and export my own logs for billing and time tracking is unacceptable. Please fix the broken API endpoint connection between the UI and the backend as soon as possible.
Attachments: (See attached screenshot of the DevTools console showing the 404 error)
