How are Google API quotas calculated?

I had a read through the following page, but wasn’t sure if this is correct information regarding Rclone and API access through it:

Just wanted to understand how the API quotas work with apps like Rclone. I read on that page that some quotas are API Console project-specific. Would this mean that you could potentially create two or three projects and load balance Rclone calls with three different credentials when it detects 403’s? Or is the requests per 100 seconds per user the one that is the main bottleneck (which is global / not project-specific).

I believe there are both limits on the project itself globally and by user. I went ahead and created my own client id just so I can have so me sort of consistent limit. good or bad…

Right, I’m just thinking if Rclone could use more than one project API credentials to swap between, to have effectively double ‘access’ to GDrive.

You can create your own and use it. Then you should only be bound by your limits.

Its not really the topic, i wanted to share my latest encounter with the googe api. While in the past I got 24h banned for scanning my library (which is obviously a google drive) todays scan didnt cause a ban!
Did anyone noticed something similar recently? Did Google change something in their quota management?
I use my own quota, not the standard shared one from rclone.