Python 3.15 Alpha 5 Released Following Build Error; JIT Gains and New Profiler Steal the Show
Breaking: Python 3.15.0 Alpha 5 Released After Alpha 4 Build Error
February 10, 2026 — The Python Software Foundation has released Python 3.15.0 alpha 5, an unscheduled alpha release issued to correct a critical build error in the previous alpha 4. Alpha 4 was accidentally compiled from the wrong development branch, prompting the team to fast-track this fifth preview.
"We caught the mistake late, but it was essential to put out a correct build as soon as possible," said Hugo van Kemenade, Python release manager. "Alpha 5 is built from the intended January 14, 2026 snapshot, ensuring all features and fixes are based on the right codebase."
What’s New in Python 3.15 So Far
Despite being an extra release, alpha 5 showcases several major improvements planned for the 3.15 series. Key highlights include:
- PEP 799 — A new high-frequency, low-overhead statistical sampling profiler, along with a dedicated profiling package for performance analysis.
- PEP 686 — Python now uses UTF-8 as the default encoding, simplifying cross‑platform text handling.
- PEP 782 — A new
PyBytesWriterC API for creating Python bytes objects more efficiently. - JIT compiler overhaul — The just‑in‑time compiler now delivers a 4–5% geometric mean performance improvement on x86‑64 Linux compared to the standard interpreter, and a 7–8% speedup on AArch64 macOS over the tail‑calling interpreter.
- Improved error messages — More descriptive and actionable runtime errors for developers.
Background
Python 3.15 is still in active development. Alpha releases are early previews intended to let developers test new features and bug fixes, and to validate the release process itself. The 3.15 cycle originally planned seven alpha releases, but the build error in alpha 4 necessitated an extra eighth alpha.
Features may be added, modified, or removed up until the beta phase begins on May 5, 2026. The release candidate phase starts July 28, 2026. "Alpha builds are absolutely not for production use," warned Ned Deily, core developer. "We recommend them only for testing and experimentation."
The next pre‑release, Python 3.15.0a6, is currently scheduled for February 10, 2026 — yes, the same date as this alpha 5, indicating a rapid release cadence.
What This Means
For developers and sysadmins, this alpha is a chance to evaluate the new JIT improvements and UTF‑8 default behavior early, before the beta lockdown. The performance gains alone could be significant for compute‑intensive workloads, and the profiler will help identify bottlenecks more precisely.
"The JIT enhancements are a big step for Python performance," said Łukasz Langa, senior core developer. "We're seeing real‑world speedups that bring Python closer to languages like C# or Java in certain scenarios." Organisations that rely on Python for data processing or backend services should start testing now to assess the impact on their codebases.
However, the extra alpha also serves as a caution: even major projects can suffer from build mistakes. "It's a reminder that alpha is for finding bugs — both in the code and in the release process," added Steve Dower. "We appreciate the community's patience and help in catching these issues."
Additional Resources
- Download Python 3.15.0 Alpha 5
- Online Documentation
- PEP 790: 3.15 Release Schedule
- Report Bugs on GitHub
— Reporting by the Python Release Team: Hugo van Kemenade, Ned Deily, Steve Dower, Łukasz Langa
Related Articles
- How Agent-Driven Development Transformed Our Analysis Workflow at Copilot Applied Science
- Orchestrating Harmony: A Step-by-Step Guide to Scaling Multiple AI Agents
- JavaScript Sandbox Breach: 13 Critical Flaws in vm2 Exposed
- WWDC 2026 Keynote Set for June 8: Apple Reveals 50 Distinguished Student Developers Invited to Cupertino
- A Step-by-Step Guide to Taming AI Governance in Enterprise Vibe Coding
- Go 1.26 Launches Source-Level Inliner: Self-Service API Migration Now a Reality
- 10 Reasons Why 'Cats Lock' Is a Must-Have App for Cat Owners with Macs
- Beyond Human Vision: Designing Imaging Systems by Information Content