Python 3.15 Arrives with Lazy Imports and Tachyon Profiling
Python 3.15 has officially launched, bringing new features like explicit lazy imports and a built-in sampling profiler. This release also marks the end of support for Python 3.10.
In short: Python 3.15 has officially launched, bringing new features like explicit lazy imports and a built-in sampling profiler. This release also marks the end of support for Python 3.10.
Software developers who rely on the Python programming language have a brand new version to explore, bringing helpful performance tools and syntax updates.
What happened, in plain words
Python 3.15 has been released, with curation by release manager Hugo van Kemenade. Software builder Simon Willison highlighted several key additions in the new version, including explicit lazy imports to speed up startup times, a new unpacking syntax for comprehensions, and a built-in sampling profiler called Tachyon. The release also means that version 3.10 is no longer supported, allowing developers to adopt newer features from Python 3.11 onward.
Key points
- Explicit lazy imports: New commands let developers load modules only when specific symbols are actually used in the code, which helps reduce slow startup times caused by heavy dependencies.
- Unpacking in comprehensions: A neat syntax addition allows unpacking inside lists, sets, and dictionaries, proposed by Adam Hartz and Erik Demaine.
- Tachyon sampling profiler: A brand new built-in sampling profiler helps developers see which functions spend the most time running by capturing stack traces from active processes.
- Performance results: Testing by Miguel Grinberg showed that Python 3.15 performs similarly to version 3.14, except for an experimental JIT feature that offered a small speed boost.
- Dropping older versions: Because 3.15 is now available, Python 3.10 is no longer supported, freeing developers to use features introduced in Python 3.11 like tomllib and datetime.UTC.
Terms explained
- Module — A file containing Python code, functions, or variables that you can reuse in other programs. Example: Like grabbing a pre-made toolkit from a shelf instead of building every tool from scratch.
- Profiler — A tool that measures how a computer program runs, tracking which parts take the most time or memory. Example: Like a stopwatch that records how many seconds you spend on each room while cleaning a house.
- JIT — Short for Just-In-Time compilation, a technique that translates code into machine instructions while the program is running to make it go faster. Example: Like a translator whispering words in your ear instantly during a live speech instead of waiting to translate the whole document later.
Why it matters
Updates to popular programming languages like Python help software creators build faster tools, manage computer memory better, and write cleaner code for applications people use every day.
What we still don't know
Performance can vary depending on the specific code and experimental settings used, and the article notes that 3.15 speeds are mostly similar to the previous version outside of experimental features.
Based on reporting from Simon Willison. This is an independent explainer, written in our own words with AI assistance; Simon Willison has not reviewed or endorsed it. Read the original for the full details.