Comment by deno

This is how it looks on the marketing pages not in real world.

For example: what is "Antigravity Entitlement"? You can buy whatever subscription (even Ultra or higher) you like and not have access to Antigravity at all. You will have to then go through some nebulous verification process which half the time is just broken. E.g. people in EU were for months blocked from verification by the Google app insisting they message an RCE number in US unreachable from EU carriers. But if you even get that far that’s actually a success. Most of the time it’s just "verification error" with no further explanation. Then you get to try to reach a human at Google. Hahaha, just kidding. That's impossible.

Then even if it works it can just log you out some day randomly and then have you go through all the verification again. Just because. Happened twice already.

Finally you have to consider what is the potential fallout of Google deciding you did some hacking or said to AI something politically incorrect and banning your entire account and all your (or even your family’s) connected accounts.

If you have anything important on Google you need to first degoogle before considering using their AI services or you're playing with fire.

The models are good but the risk to benefit ratio is terrible.