Astral benchmarks show uv's CPython 3.14 builds beat Docker and conda-forge Python
charliermarsh · x · 2026-09-11
Astral founder Charlie Marsh published new benchmark data for python-build-standalone: on pyperformance, its CPython 3.14.6 builds compare favorably against alternative distributions.
- Methodology and raw data are published in BENCHMARKS.md on GitHub, using CPython 3.14.6
- Distributions compared include the official Docker python:3.14 image and conda-forge Python 3.14.6
- Violin plots show per-benchmark runtime ratios on a log scale; ratios above 1 mean python-build-standalone is faster, with geometric means labeled per distribution
- Marsh frames this as a long-standing goal: installing CPython via uv should give you the fastest available build by default
Related event: Astral Benchmarks python-build-standalone as Fastest CPython Distribution(4 posts)→
More from coding & agent
- Seroter Daily #864: behavioral evals for coding agents, Skills CLI 1.0, AI cost tools — rseroter · 2026-09-11
- Notion launches MCP so your memory follows you across tools, models and agents — _AustinCalvert_ · 2026-09-11
- Scott Belsky: trust, progressive personalization and agent-to-agent effects define the 'first mile' — _AustinCalvert_ · 2026-09-11
- Pydantic's Monty goes commercial as banks quietly spend millions building their own sandboxes — samuelcolvin · 2026-09-11
- Early take: DeepSeek V4.1 Solo beats Agent Teams and GLM 5.3 Flash on quality and cost — teortaxesTex · 2026-09-11
- AI movie studio builds 3D camera previs tool to save generation credits — purecharisma2020 · 2026-09-11