All these are fully implemented and verified in the typing module.
For decades, the Global Interpreter Lock (GIL) limited Python execution to a single CPU core per interpreter instance, crippling CPU-bound multi-threading. Driven by PEP 703 , Python 3.13 introduces an experimental build-time configuration to fully disable the GIL. What's New In Python 3.13 — Python 3.14.5 documentation
: Multi-line snippets containing uneven indentation or mixed statements can now be pasted directly into the prompt without triggering standard indentation errors. python 313 release notes verified
: Multi-threaded programs can now scale workloads across multiple CPU cores natively.
# The new REPL beautifully preserves and handles multi-line blocks seamlessly: def calculate_growth(value): if value > 0: return value * 1.13 return 0 Use code with caution. All these are fully implemented and verified in
if:
Verified: A copy-and-patch JIT compiler is added behind a build flag ( --enable-experimental-jit ). Reality: Not a speed miracle yet. It translates bytecode to machine code at runtime, but initial benchmarks show modest gains (5–15% in some loops). The foundation is laid — expect major improvements in 3.14. What's New In Python 3
The most impactful updates are currently experimental and require specific builds or flags to enable.
: The locals() built-in function now produces a clearly defined snapshot. This allows external software debuggers to reliably mutate local variables within optimized scopes during execution.
: While the current performance uplift provides a modest baseline for standard scripts, this JIT architecture serves as the foundation for massive performance optimizations across upcoming Python cycles. 3. A Completely Revamped Interactive REPL
If you are stuck on an ancient Linux distribution (e.g., CentOS 7 with OpenSSL 1.0.2), Python 3.13 will not compile or run without backporting a newer OpenSSL.