With the release of GPT-4o-mini, I’m updating some of the models used in Micro.blog. It’s really nice to see the price decrease. Hopefully that is translating to more efficient, less energy use on the servers too.