Now they've left a hybrid Flash model that doesn't work....
I don't know anything about coding or programming, but I do know about Antigravity. I've practically used it 12 to 15 hours a day for 4 months. What I see is that with the recent change, they had put in a 3.5 Flash model that burned through tokens brutally, but it completed tasks in seconds and without errors. However, the quota only reset after 6 days, so I canceled my Pro plan, and I can use it until May 24th.
Now, seeing that this model barely left any room for usage, they changed it for another one, claiming they 'increased' the limits. And yes, now the model resets every 4 hours, but it performs poorly. I've gone through 4 token resets since yesterday, and it can't even finish a simple app translation task. It constantly repeats the same processes and keeps saying it's about to proceed, but it burns through all the credit in pure analysis and review scripts.
