Is switching Google Antigravity accounts a ToS violation?

Google Antigravity
1

Hey everyone, I have two active paid accounts on Google Antigravity, and can I switch between them to manage my token limits. Is this considered a violation of the terms of service?

2

I did it few times before i moved to other IDE, and still my accounts are live.

3

Hi @DasikIgaijinn,

While we cannot share the exact details of our abuse detection protocol, switching between two active, paid accounts to manage token usage is not considered a violation of our Terms of Service.

However, please keep in mind that issues often arise when users attempt to use secondary accounts that are part of a discounted or shared pool. To ensure uninterrupted access and avoid triggering any system flags, please make sure you are only using paid accounts that you personally own.

4

Hello,

I am an individual developer with 2 Google accounts, all mine, each with its own [Google AI Pro / Ultra] subscription that includes Antigravity. Before I go further, I want to make sure my setup is within the Antigravity terms.

My setup: a personal tool that runs only on my own computer, used only by me, for my own software work. It signs in with Google (the Antigravity sign-in) and sends requests to Antigravity. I do not share it, resell access or let anyone else use it.

My questions:

  1. Is it allowed to use my Antigravity subscription from a personal tool like this, instead of the Antigravity IDE or the agy CLI? If not, which official option do you recommend for this use (for example the Gemini API with a paid key)?
  2. When one of my accounts reaches its usage limit, may I continue on another of my own subscribed accounts? Is this allowed if I switch manually? Is it allowed if my tool switches automatically?
  3. Large requests (a prompt of about 100 KB with about 25 tool definitions) get HTTP 429 RESOURCE_EXHAUSTED (“Resource has been exhausted (e.g. check quota)”) immediately, even though the quota shown for the account is at 100%. Smaller requests on the same account and model succeed. Is there a per-request or per-minute limit I should respect?

I would appreciate a written answer, so that I can configure my tool correctly.

Thank you,
Adrian