You can run litellm as a proxy to not even worry about changes in the API (much).

It still makes sense for very small teams where the burden of maintaining the infra is too high, but if you have 2k engineers it’s a no brainer to self host. It’s only a matter of time before open weight models can process vision as well.

I’ve begun switching from jetbrains IDEs to Zed because of the memory usage. I want snappy.

You can’t have our mayor. Out of our cold dead hands