The first question is always: which model? Which fortunately you at least addressed: >free plan of ChatGPT, but I can remember if it was 4o or 4. Since chatgp

The first question is always: which model? Which fortunately you at least addressed: >free plan of ChatGPT, but I can remember if it was 4o or 4. Since chatgpt-4o, there has been o1-preview, and o1 (full) is out. They just announced o3 got a 25% on frontiermath which is what this article is a reaction to. So, any tests on 4o are at least TWO (or three) AI releases with new capabilities.