
The Real Python Podcast Benchmarking Python 3.14 & Enabling Asyncio to Scale
23 snips
Oct 24, 2025 Join Christopher Trudeau, a seasoned Python educator and contributor, as he dives into the latest benchmarks for Python 3.14. He reveals how asyncio performs with the new free-threaded build and discusses several impactful PEPs that aim to enhance Python's packaging and async capabilities. Explore articles on library compatibility, the concept of good taste in software engineering, and a novel tool for testing documentation. Plus, discover how to build reactive web applications using NiceGUI, all while diving deep into the vibrant Python community!
AI Snips
Chapters
Books
Transcript
Episode notes
3.14 Restores Notable Single-Thread Speed
- Python 3.14 shows measurable single-thread speed improvements over 3.13 in hand-crafted benchmarks.
- The release returns performance gains started in 3.11 and continues optimization across versions.
Free‑Threading Helps With CPU‑Hungry Threads
- Free-threaded mode (GIL removed) shows limited single-thread gains but shines with CPU-bound multi-threaded workloads.
- Combining free-threaded mode with the JIT produces the biggest multi-thread performance wins.
Verify Libraries For Free‑Threaded Mode
- Check library compatibility before switching to free-threaded Python using ft-checker and the PSF heatmap.
- Search specific packages to see if they run in free-threaded mode before upgrading production environments.



