Note that this is RC3; 3.14.0 final will be out soon, but not yet.
A lot of people have been looking forward to this simply because of the pi memes. Personally I'm most excited about the PEP 750 template strings and PEP 784 zstd compression in the standard library. (Although it would be nice to have a higher-level interface to the compression standard libraries in general — I might end up designing something....) And of course, error messages and the REPL continue to improve and I'm quite happy for that. But people who were expecting big performance gains from the new work on a free-threaded interpreter, experimental JIT etc. are going to have to wait.
A lot of people have been looking forward to this simply because of the pi memes. Personally I'm most excited about the PEP 750 template strings and PEP 784 zstd compression in the standard library. (Although it would be nice to have a higher-level interface to the compression standard libraries in general — I might end up designing something....) And of course, error messages and the REPL continue to improve and I'm quite happy for that. But people who were expecting big performance gains from the new work on a free-threaded interpreter, experimental JIT etc. are going to have to wait.
I'll also take a moment to note that the naming of https://docs.python.org/3.14/reference/datamodel.html#object... was my suggestion (https://zahlman.github.io/posts/2025/01/11/a-brief-annotatio...) — not that I can ever expect the credit to be given officially. (And frankly, it's almost too obvious to be worth crediting; yet somehow I had to think of it for it to happen.) The corresponding PEP (https://peps.python.org/pep-0649/) is now well over four years old, having been accepted two years ago; there has been support for the idea, but it hasn't been clear how to minimize disruption while releasing it (hence the separate https://peps.python.org/pep-0749/).