> which is the standard for all of Google's public APIs
Far from that, unfortunately. A lot of Google public APIs predate AIPs and it's practically impossible to fix them and make them compliant. Possibly the worst offender is the Compute Engine API [1], which, to the best of my knowledge, is REST only (no gRPC), uses its own conventions related to page size / next page token, and, in general, is special in many ways. One of the ways it's special is its size: its definition in Google's own Discovery format [2] is a 6MB JSON.
It's an extremely difficult problem to standardize multiple APIs produced by multiple teams, and then keep the new changes compliant. Google did a great job, but it's still very far from the perfect state.
I worked on this for a couple years (have been much less active lately), surprised to see it on the front page of HN but happy to answer questions about it.
More context at [2].
[1]: https://aip.dev [2]: https://aep.dev/blog/history-of-aeps/
Far from that, unfortunately. A lot of Google public APIs predate AIPs and it's practically impossible to fix them and make them compliant. Possibly the worst offender is the Compute Engine API [1], which, to the best of my knowledge, is REST only (no gRPC), uses its own conventions related to page size / next page token, and, in general, is special in many ways. One of the ways it's special is its size: its definition in Google's own Discovery format [2] is a 6MB JSON.
It's an extremely difficult problem to standardize multiple APIs produced by multiple teams, and then keep the new changes compliant. Google did a great job, but it's still very far from the perfect state.
[1]: https://docs.cloud.google.com/compute/docs/reference/rest/v1 [2]: https://www.googleapis.com/discovery/v1/apis/compute/v1/rest - warning: 6 megabytes JSON
[edit: typo]
But yes, Google has plenty of legacy APIs. And yes, GCE in particular is a HUGE pain.
It has been super helpful and the ecosystem is growing around it.
The real benefit is having a standardized pattern for APIs, with automation and predictability built in.
Not everything can be modeled as a resource, but for any API that fits that, it’s fantastic.
It’s a bit restrictive at times, but those restrictions often shed light on bad architecture/data modeling.
They worked through a lot of feedback last year and cut a 2026 LTS release to give strong confidence to start building tooling around it.
Check out https://aepbase.io/ for an example and fun way to get started.