Python 3.15 is still in active development, and the sixth alpha release—3.15.0a6—has just landed. This early developer preview gives us a sneak peek at the major changes coming in the final 3.15 series. While you shouldn’t use it in production (alpha releases are unstable by design), it’s the perfect opportunity for testing and experimentation. Below, we break down the eight most important things you need to know about this release, from new PEPs to performance boosts. Let’s dive in.
1. What Is Python 3.15.0 Alpha 6?
This is the sixth of eight planned alpha releases for Python 3.15. It’s designed to give developers early access to new features and bug fixes, and to help the core team validate the release process. The alpha phase runs until May 5, 2026, after which the beta phase begins. Features can still be added, modified, or even removed until the release candidate phase on July 28, 2026. For now, treat this as a sandbox—great for testing, but not for production workloads.
2. PEP 799: A New High-Frequency Profiler
One of the most anticipated additions is PEP 799, which introduces a statistical sampling profiler designed for low overhead and high frequency. This profiler lives in its own dedicated package, making it easy to integrate into your performance analysis workflow. Unlike traditional profilers that instrument every function call, this one samples the call stack at regular intervals, providing a truer picture of where your code spends its time—without dragging down performance.
3. PEP 798: Unpacking in Comprehensions
PEP 798 brings a long-requested feature: the ability to unpack iterables inside list, dict, and set comprehensions using * and ** operators. For example, you can now write [*(a, b) for a, b in pairs] or {**d for d in dicts}. This makes it easier to flatten nested structures or merge dictionaries inline, reducing boilerplate code and improving readability.
4. PEP 686: UTF-8 as the Default Encoding
After years of debate, Python 3.15 makes UTF-8 the default text encoding for open(), input(), and other I/O operations. This aligns Python with modern best practices and eliminates many cross-platform encoding bugs. If you’ve been bitten by locale-dependent defaults in the past, this change will come as a relief. Note that you can still override the encoding manually if needed.
5. PEP 782: PyBytesWriter C API
For extension writers and performance enthusiasts, PEP 782 introduces a new C API called PyBytesWriter. This allows you to efficiently build Python bytes objects from C code without the overhead of repeated allocations. It’s especially useful when constructing large binary payloads or implementing custom serializers. Look for this API in the C extensions docs.
6. PEP 728: TypedDict with Extra Items
Type hints get a boost with PEP 728, which extends TypedDict to allow specifying typed extra items. You can now define a TypedDict that accepts any extra keys, but with a constraint on their values’ types. This bridges the gap between strict typing and real-world flexible dictionaries, making it easier to model JSON data or configuration files.
7. JIT Compiler Performance Gains
The JIT compiler—introduced in earlier Python versions—has been significantly upgraded. Benchmarks show a 3–4% geometric mean improvement on x86-64 Linux over the standard interpreter, and an even more impressive 7–8% speedup on AArch64 macOS compared to the tail-calling interpreter. These gains come from better code generation and optimization passes, making Python faster without changing your code.
8. Improved Error Messages and What’s Next
As with every recent Python release, error messages have been refined to be more helpful. You’ll see clearer hints for syntax errors, better suggestions for attribute typos, and improved tracebacks. The next pre-release, 3.15.0a7, is scheduled for March 10, 2026. Keep an eye on the release schedule and report issues on the CPython issue tracker.
Python 3.15 is shaping up to be a solid release with practical improvements for developers and users alike. Whether you’re excited about the new profiler, the JIT speedups, or the cleaner encoding defaults, there’s something here for everyone. Download the alpha from python.org, give it a spin, and help shape the final release. Happy coding!