I am trying to run Text-to-Speech in my production app (hosted on another cloud; no workload identity federation), and I'm trying to use IAM as recommended by the docs.
Specifically, I'm hitting the text.synthesize method.
The Text-to-Speech IAM page says:
Note: The IAM permissions listed on this page are for Long Audio Synthesis API, and don't apply to the original Synthesis API.
So, are there IAM roles for the original Synthesis API?
I've looked through the IAM lists in the console many times in different UIs and have never found it. Right now I'm just using a service account with universal access to Google Cloud. This is strongly recommended against in the docs, but I don't see any alternative.