PyCon Sweden 2024

Articles

The State of Python 2025

Michael Kennedy reviewed over 30,000 responses from the eighth annual Python Developers Survey and shared the key trends, insights and takeaways in this post.

# Michael Kennedy

How to Use Redis with Python

Many of us use Redis strings for caching, but Redis actually supports several other data types like Hashes, Lists, Sets, Bitmaps, HyperLogLogs, etc. Each of these comes with its own commands: for example, HSET, HGET and HDEL for hashes, or LPUSH, LGET and LREM for lists. In this article, Federico Trotta explained some of these data types and showed how to use them with examples.

# Federico Trotta

Custom parametrization scheme with Pytest

Parametrizing tests makes it easy to run them across multiple scenarios. But the default syntax has one drawback, it’s hard to tell which parameters belong to which scenario just by looking at the logs. In this article, Christos Liontos shared a clever solution using a custom hook to tackle this problem.

# Christos Liontos

How Well Do New Python Type Checkers Conform? A Deep Dive into Ty, Pyrefly, and Zuban

Python’s type checking space has been very active recently. Three new tools - Ty, Pyrefly and Zuban, have emerged as promising options. In this article, Rob Hand reviewed the popular type checkers, shared highlights of these new tools, shared Conformance Suite analysis results of them and shared the gap between conformance and real-world usage.

# Rob Hand

Production-Grade Python Logging Made Easier with Loguru

Loguru is one of popular logging libraries, many say it’s what Python’s default logging should have been. If you’d like to learn how to use it, check out this article by Ayooluwa Isaiah. He covered everything from Loguru’s philosophy to how you can bring your Python logs into an observability pipeline.

# Ayooluwa Isaiah

Interesting Packages and Projects to Explore

Thonny - Python IDE for beginners

Cosmic Ray - Mutation testing

Tach - A Python tool to maintain a modular package architecture.

Expression - Practical functional programming for Python 3.10+

TkForge - Drag & drop in Figma to create a Python GUI with ease

Recent Noteworthy Package Releases

pydantic-ai 1.0.0 - Agent Framework / shim to use Pydantic with LLMs

haystack-experimental 0.13.0 - Experimental components and features for the Haystack LLM framework

lancedb 0.25.0 - Developer-friendly, embedded retrieval engine for multimodal AI. Search More; Manage Less

graphiti-core 0.20.x - A temporal graph building library

torchao 0.13.0 - Package for applying ao techniques to GPU models

dag-factory 1.0.0 - Dynamically build Apache Airflow DAGs from YAML files

certbot-dns-cloudflare 5.0.0 - Cloudflare DNS Authenticator plugin for Certbot

pyrefly 0.31.0 - A fast Python type checker written in Rust

fastexcel 0.15.x - A fast excel file reader for Python, written in Rust

pyexasol 1.0.0 - Exasol python driver with extra features

trl 0.22.x - Train transformer language models with reinforcement learning

inference-cli 0.54.x - With no prior knowledge of machine learning or device-specific deployment, you can deploy a computer vision model to a range of devices and environments using Roboflow Inference

sparkmeasure 0.26.0 - Python API for sparkMeasure, a tool for performance troubleshooting of Apache Spark workloads

tokenizers 0.22.0 - 💥 Fast State-of-the-Art Tokenizers optimized for Research and Production

Want More? Click Here

Python: The Documentary | An origin story

pypistats.org is now operated by the Python Software Foundation

Articles

Starting with pytest’s parametrize

If you’re writing test cases, I strongly suggest using pytest’s parametrize feature. It helps you save both time and code. In this article, Ned Batchelder explained how to use it with a clear example.

# Ned Batchelder

Litestar is worth a look

Litestar is a Python web framework that’s built with async support and type hints in mind. You might think, “But we already have FastAPI.” That’s true, but James Bennett shared some good reasons why this framework is worth exploring.

# James Bennett

Proper Guided Roadmap to learn the Python programming language

A complete guide to learning Python from beginner to advanced, created by Santosh Bhandari.

# Santosh Bhandari

Favourite Python tricks or lesser-known features

In this Reddit post, people shared some really cool Python features with examples like itertools.batched, the walrus operator, python -m http.server, defaultdict, functools.partial and more.

Expressions are coming to pandas!

Pandas was introduced 17 years ago and quickly became the go-to tool for data science in Python. Over time, new libraries appeared, building on pandas’ ideas while fixing some of its limitations. Now, pandas has come full circle with a new syntax inspired by those newer libraries. In this article, Marco Gorelli explains why this change happened and how you can use it.

# Marco Gorelli

Interesting Packages and Projects to Explore

FastAPI boilerplate - An extendable async API using FastAPI, Pydantic V2, SQLAlchemy 2.0, PostgreSQL and Redis

mutmut - Mutation testing for Python

cruft - Allows you to maintain all the necessary cruft for packaging and building projects separate from the code you intentionally write. Built on-top of CookieCutter

pyee - A rough port of Node.js's EventEmitter to Python with a few tricks of its own

Pyxel - A retro game engine for Python

Recent Noteworthy Package Releases

protovalidate 0.15.0 - Protocol Buffer Validation for Python

onnx 1.19.0 - Open Neural Network Exchange

cron-descriptor 2.0.x - A Python library that converts cron expressions into human readable strings

rpaframework 31.0.0 - A collection of tools and libraries for RPA

litellm 1.76.0 - Library to easily interface with LLM API providers

onnxscript 0.4.0 - Naturally author ONNX functions and models using a subset of Python

Want More? Click Here

Preventing ZIP parser confusion attacks on Python package installers

Python 3.14.0rc2 released

The next pre-release of Python 3.14 will be the final release candidate, 3.14.0rc3, scheduled for 2025-09-16; the official release of 3.14.0 is scheduled for Tuesday, 2025-10-07

Python 3.13.7 released

This is the seventh maintenance release of Python 3.13

Nominees for 2025 PSF Board Election Announced

The last supported Python version for Pytype will be 3.12

PyData Seattle CFP is open

[Accepted] PEP 728 – TypedDict with Typed Extra Items

[Final] PEP 750 – Template Strings

[Final] PEP 792 – Project status markers in the simple index

PyPI now checks for expired domains to prevent domain resurrection attacks

Articles

The Recursive Leap of Faith, Explained

Al Sweigart wrote an amazing article on writing recursive functions, breaking it down into five simple guidelines. If recursion feels tricky, his approach makes it much easier to understand and apply in real-world problems.

# Al Sweigart

4 Text Similarity Tools: When Regex Isn't Enough

Khuyen Tran wrote a nice article on text similarity in Python. She covered multiple approaches: Regex, difflib, RapidFuzz, and Sentence Transformers. She shared examples for each and explained when to use which approach.

# Khuyen Tran

Checking for string prefixes and suffixes in Python

Two often-overlooked Python string methods are startswith() and removeprefix(). Trey Hunner covered them with nice examples. Remembering these can help you write cleaner and more readable code.

# Trey Hunner

The Python package ecosystem is huge and sometimes it’s tough to discover the right tools when you need them. That’s why it’s always a good idea to stay updated on useful packages. In this video, Arjan Egges covered 15 powerful but often overlooked packages that can help you write cleaner code and boost performance, with practical real-world examples.

# Arjan Egges

Exciting 40+ Projects on Deep Learning to Enhance Your Portfolio in 2025

Many freshers often end up building the same set of projects, which doesn’t always help them stand out. If you’re a fresher and want to try something different to test and showcase your skills, check out this article by Kechit Goyal. He shared 40+ project ideas that can help you learn, practice and build a stronger portfolio.

# Kechit Goyal

Interesting Packages and Projects to Explore

FinanceToolkit - Transparent and Efficient Financial Analysis

xonsh - Python-powered shell. Full-featured and cross-platform

Cappa - Declarative CLI argument parser

Awesome Public Datasets - A topic-centric list of HQ open datasets

yt-dlp - A feature-rich command-line audio/video downloader

Recent Noteworthy Package Releases

fastapi-pagination 0.14.0 - FastAPI pagination

compressed-tensors 0.11.0 - Library for utilization of compressed safetensors of neural network models

browser-use 0.6.0 - Make websites accessible for AI agents

mlflow 3.3.0 - MLflow is an open source platform for the complete machine learning lifecycle

langfuse 3.100.0 - A client library for accessing langfuse

diffusers 0.35.0 - State-of-the-art diffusion in PyTorch and JAX

ntc-templates 8.0.0 - TextFSM Templates for Network Devices, and Python wrapper for TextFSM's CliTable

zstandard 0.24.0 - Zstandard bindings for Python

pint 0.25.0 - Physical quantities module

dspy 3.0.0 - The framework for programming—not prompting—language models

markdown-it-py 4.0.0 - Markdown parser, done right. 100% CommonMark support, extensions, syntax plugins & high speed

Want More? Click Here

The PSF has paused our Grants Program

The PSF has temporarily paused its Grants Program after hitting the 2025 budget cap earlier than expected.

PyCon UK 2025 Schedule

Articles

2048: iterators and iterables

Iterable and iterator are not the same, though they’re often confused. In this article, Ned Batchelder explained the difference with a simple example and a great analogy: the pages of a book are an iterable; a bookmark is an iterator.

# Ned Batchelder

Don't call dunder methods

Dunder methods (like __init__, __str__, etc.) are low-level hooks used by the Python interpreter to enable certain behaviours. While they’re powerful, using them directly is often unnecessary and can lead to confusing code. In this article, Trey Hunner explained why it’s best to avoid them unless absolutely needed, with clear examples and shared alternatives.

# Trey Hunner

Why Python’s deepcopy Can Be So Slow (and How to Avoid It)

Python’s copy.deepcopy() is powerful but very slow and memory-heavy. If you're considering using it, check out this article by Saurabh Misra. He explained why it's often overkill and shared alternatives - like preferring shallow copies, copying only what's needed and other tips to make your code faster and more efficient.

# Saurabh Misra

Using Python Pandas to Clean and Analyze Scraped Data

If you work with data, you probably know that 95% of the job is cleaning and prepping it. In this article, Hassan Rehan highlighted the importance of data cleaning and showed how to use Pandas for processing, along with key techniques to clean and analyse scraped data effectively.

# Hassan Rehan

Agents Towards Production

Nir Diamant’s GitHub repository features 30+ in-depth tutorials for building AI agents. He covered everything from stateful workflows, vector memory and real-time web search APIs to Docker deployment, FastAPI endpoints, security guardrails, GPU scaling, browser automation, fine-tuning, multi-agent coordination, observability, evaluation and even UI development. A good resource for anyone building serious AI systems.

# Nir Diamant

Interesting Packages and Projects to Explore

dynaconf - The dynamic configurator for your Python Project

Pyro5 - Remote object communication library

Rio - Build modern Websites and Apps just with Python

PyCryptodome - Cryptographic library for Python

YTSage - Modern YouTube downloader with a clean PySide6 interface

Recent Noteworthy Package Releases

cuda-bindings 13.0.0 - Python bindings for CUDA

pydantic-ai 0.6.0 - Agent Framework / shim to use Pydantic with LLMs

torchvision 0.23.0 - image and video datasets and models for torch deep learning

curl-cffi 0.13.0 - libcurl ffi bindings for Python, with impersonation support

trl 0.21.0 - Train transformer language models with reinforcement learning

pyvista 0.46.0 - Easier Pythonic interface to VTK

protovalidate 0.14.0 - Protocol Buffer Validation for Python

mlflow 3.2.0 - An open source platform for the complete machine learning lifecycle

redis 6.3.0 - Python client for Redis database and key-value store

peft 0.17.0 - Parameter-Efficient Fine-Tuning (PEFT)

cartopy 0.25.0 - A Python library for cartographic visualizations with Matplotlib

Want More? Click Here

PyPI Users Email Phishing Attack

If you’ve published projects on PyPI and included your email in the package metadata, watch out for phishing emails titled "[PyPI] Email verification" coming from noreply@pypj.org (note the lowercase j in the domain, not the official PyPI domain). Mike Fiedler has shared a preliminary report on this issue. Stay cautious!

# Mike Fiedler

Articles

Building Autonomous Systems: A Guide to Agentic AI Workflows

Agentic AI Workflows - systems that can reason, plan and act autonomously are becoming a hot topic. Unlike traditional hardcoded orchestration, these workflows operate across multiple steps, tools and platforms. If you want a solid introduction, check out this article by Adrien Payong and Shaoni Mukherjee, where they explained the core concepts, tools and frameworks behind it.

# Adrien Payong & Shaoni Mukherjee

Python Gotcha: Logging an uncaught exception

Unhandled exceptions can be confusing, especially when there's no clear stack trace. Andrew Wegner shared a neat solution using sys.excepthook to make debugging easier and more transparent.

# Andrew Wegner

10 Surprising Things You Can Do with Python’s collections Module

The collections module offers powerful alternatives to Python's built-in containers like dict, list, set, and tuple. These can simplify code, improve readability and even boost performance. In this article, Matthew Mayo shared 10 amazing examples to help you understand how to use them effectively.

# Matthew Mayo

Tim Ruscica shared an amazing tutorial on building MCP servers in Python by creating a simple MCP server for user notes access. He went beyond the basics and used Stytch for handling authorization instead of traditional access tokens.

# Tim Ruscica

How to Debug Common Python Errors

If you’re not comfortable reading tracebacks or don’t know how to debug, building solid software becomes tough. In this beginner-friendly article, Theresa Obamwonyi explained the basics of tracebacks and debugging - perfect for anyone new to this topic.

# Theresa Obamwonyi

Interesting Packages and Projects to Explore

marimo - A library for making reactive notebooks and apps

Toolz - List processing tools and functional utilities

Uplink - A Declarative HTTP Client for Python

chameleon-flask - Adds integration of the Chameleon template language to Flask and Quart

Diffusers - State-of-the-art diffusion in PyTorch and JAX

Recent Noteworthy Package Releases

llama-index 0.13.0 - Interface between LLMs and your data

pyOCD 0.38.0 - Cortex-M debugger for Python

pip 25.2.0 - The Python package installer

optype 0.13.0 - Building blocks for precise & flexible type hints

elasticsearch 9.1.0 - Python client for Elasticsearch

Keras 3.11.0 - Deep Learning for humans

flax 0.11.0 - A neural network library for JAX designed for flexibility

docutils 0.22.0 - Python Documentation Utilities

trl 0.20.0 - Train transformer language models with reinforcement learning

langgraph 0.6.0 - Building stateful, multi-actor applications with LLMs

narwhals 2.0.0 - Extremely lightweight compatibility layer between dataframe libraries

autogen 0.7.1 - A programming framework for agentic AI

sse-starlette 3.0.0 - SSE plugin for Starlette

lightning-utilities 0.15.0 - Lightning toolbox for across the our ecosystem

optree 0.17.0 - Optimized PyTree Utilities

Want More? Click Here

Run marimo in the cloud

Marimo team announced in their blog post about molab, it is a cloud-hosted marimo notebook workspace that lets us rapidly experiment on data using Python and SQL.

Python 3.14 release candidate 1 is released!

Articles

Django's Ecosystem

A programming language or framework becomes popular when it offers numerous helpful tools and resources. Django is one of the great examples of it. The Django Software Foundation has put together a useful list with information, including packages for debugging, API development and more.

# Django Software Foundation

Python Backstage • Disassembling Python Code Using the `dis` Module

We all know that the Python interpreter changes our code into an intermediate form called bytecode. With the dis module, you can see what this bytecode looks like. In this article, Stephen Gruppetta explained what happens inside when we run a Python program.

# Stephen Gruppetta

Run coverage on tests

Code coverage tools show which parts of your program were run during testing. Do you know they are also useful for other things, like finding tests that didn’t run? In this post, Hugo van Kemenade explained why we should use a coverage tool, with two easy examples.

# Hugo van Kemenade

How good is your knowledge of Python f-strings?

If you think you know a lot about f-strings, try this quiz made by Armin Ronacher. I scored 11 and learned that you can do many cool things with f-strings. I hope you have fun & learning too!

# Armin Ronacher

Scaling recommendations service at OLX

In distributed systems, there is a saying: “You are as slow as your slowest task.” If you don’t make your slowest part faster, the whole system will be as slow as that part. In this article, Jordi Esteve Sorribas explained the changes they made to a FastAPI service to help it handle tens of thousands of requests every second, with almost all responses happening in under 10 milliseconds.

# Jordi Esteve Sorribas

Interesting Packages and Projects to Explore

jq.py - a lightweight and flexible JSON processor

Arakawa - Build interactive reports in seconds using Python

pydantic-cli - Turn Pydantic defined Data Models into CLI Tools

PyTestArch - Test framework for software architecture based on imports between modules

PyHamcrest - Hamcrest framework for matcher objects

Recent Noteworthy Package Releases

Textual 5.0.0 - Modern Text User Interface framework

huggingface_hub 0.34.0 - Client library to download and publish models, datasets and other repos on the huggingface.co hub

vLLM 0.10.0 - A high-throughput and memory-efficient inference and serving engine for LLMs

line_profiler 5.0.0 - Line-by-line profiler

OpenCLIP 3.0.0 - Open reproduction of consastive language-image pretraining (CLIP) and related

CrewAI 0.150.0 - Cutting-edge framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks

cmaes 0.12.0 - Lightweight Covariance Matrix Adaptation Evolution Strategy (CMA-ES) implementation for Python 3

JAX 0.7.0 - Differentiate, compile, and transform Numpy code

Kedro 1.0.0 - Kedro helps you build production-ready data and analytics pipelines

Logfire 4.0.0 - The best Python observability tool! 🪵🔥

Feast 0.51.0 - Python SDK for Feast

timezonefinder 7.0.0 - python package for finding the timezone of any point on earth (coordinates) offline

GeoAlchemy2 0.18.0 - Using SQLAlchemy with Spatial Databases

Want More? Click Here

PyCon Greece 2025 – Schedule is Live

PEP 734 – Multiple Interpreters in the Stdlib (Final)

PEP 792 – Project status markers in the simple index (Accepted)

Articles

Hynek Schlawack released another great video on uv, where he explained how he uses the just tool to store commands in a cross‑platform, portable way for everyday tasks like installing or refreshing virtual environments, running tests and code checks and even development tasks like sending requests.

# Hynek Schlawack

From pip to a uv project

The Astral team published a guide on moving from a pip and pip‑tools workflow based on requirements files to uv’s project workflow using pyproject.toml and uv.lock.

# Astral

Use TorchAudio to Prepare Audio Data for Deep Learning

Ever wondered how machine learning models process audio data and handle varying audio lengths while staying robust?
If yes, check out this tutorial by Negar Vahid, where she walked through handling audio data with TorchAudio, a PyTorch‑based toolkit.

# Negar Vahid

Building a Multi-tenant App with Django

Multitenancy is a key aspect to consider when building a SaaS application. In this article, Nik Tomazic explained three multi‑tenancy architecture strategies and demonstrated how to implement one using Django, django‑tenants and django‑tenant‑users.

# Nik Tomazic

Audio, Video, and Webcams in Python (Lost Chapter from Automate the Boring Stuff)

Python makes it easy to automate all kinds of tasks—one of the big reasons so many of us love it. In this article, Al Sweigart demonstrates how to automate various tasks involving audio, video, and webcams with clear examples.

# Al Sweigart

Interesting Packages and Projects to Explore

FastAPI Model Server Skeleton - FastAPI Skeleton App to serve machine learning models production-ready

ZipNN - A Lossless Compression Library for AI pipelines

Tweepy - Library for accessing the X API (Twitter)

Runtype - Type dispatch and validation for run-time Python

deepface - A Lightweight Face Recognition and Facial Attribute Analysis Framework (Age, Gender, Emotion, Race) for Python

Recent Noteworthy Package Releases

uv 0.8.0 - An extremely fast Python package and project manager, written in Rust.

TorchAO 0.12.0 - Package for applying ao techniques to GPU models

Firebase Admin Python SDK 7.0.0

ragas 0.3.0 - Supercharge Your LLM Application Evaluations 🚀

coredis 5.0.0 - Python async client for Redis key-value store

Streamlit 1.47.0 - A faster way to build and share data apps

supervision 0.26.0 - A set of easy-to-use utils that will come in handy in any Computer Vision project

dag-factory 0.23.0 - Dynamically build Apache Airflow DAGs from YAML files

kedro-datasets 8.0.0 - Kedro-Datasets is where you can find all of Kedro's data connectors.

cyclonedx-python 7.0.0 - CycloneDX Software Bill of Materials (SBOM) generator for Python projects and environments

Crawl4AI 0.7.0 - Open-source LLM Friendly Web Crawler & scraper

textual 4.0.0 - Modern Text User Interface framework

opentelemetry-python 1.35.0 - OpenTelemetry Python API

Want More? Click Here

Thinking about running for the PSF Board?

PyData London 2025 Videos

Python 3.14.0 beta 4 is here!

This release, 3.14.0b4, is the last of four planned beta releases.

In Python 3.14, Free-threaded will be fully supported as non-experimental

Articles

Koan 1: The Empty Path

In Python, writing if some_object: is super common and convenient, but have you ever wondered how it actually works behind the scenes or how you can customise that behaviour for your own classes? If not, then you should definitely check out this article by Vivis Dev, where both questions are answered clearly with practical examples.

# Vivis Dev

Making a Simple HTTP Server with Asyncio Protocols

In this article, Jacob Padilla walked through building a simple HTTP server from scratch using asyncio Protocols - no external libraries. It's a great way to understand how an HTTP server works under the hood.

# Jacob Padilla

Building and Sharing a Practical Python Security Checklist

Some Python functions and tools can introduce subtle security risks if not used carefully. In this article, Adrien highlights examples like eval, exec, pickle, XSS issues and insecure password hashing, along with safer alternatives to use instead.

# Adrien

Application Logging in Python: Recipes for Observability

Logs are crucial for understanding what's happening inside your application, especially in production, where debugging isn't an option. Along with metrics and traces, effective logging provides a clear picture of issues and behaviour. This article by the dash0 team is a must-read for mastering logging in Python. It covers structured JSON logs, centralised logging config and enriching logs with contextvars and other modern observability practices.

# dash0 team

Simon Willison shared a short demo where he used Claude Code to implement a new GitHub Actions workflow to add a feature to one of his GitHub repositories. A neat example of how AI can assist in real development tasks.

# Simon Willison

Interesting Packages and Projects to Explore

Injector - Python dependency injection framework, inspired by Guice

Ibis - The portable Python dataframe library

Pegen - CPython's PEG parser generator

htmy - Async, pure-Python server-side rendering engine

cryptography - cryptography is a package which provides cryptographic recipes and primitives to Python developers

Recent Noteworthy Package Releases

pybind11 3.0.0 - Seamless operability between C++11 and Python

mcp 1.11.0 - Model Context Protocol SDK

great-tables 0.18.0 - Easily generate information-rich, publication-quality tables from Python.

google-adk 1.6.1 - Agent Development Kit

Datasets 4.0.0 - HuggingFace community-driven open-source library of datasets

Pandera 0.25.0 - A light-weight and flexible data validation and testing tool for statistical data objects.

browser-use 0.5.x - Make websites accessible for AI agents

SQLGlot 27.0.0 - An easily customizable SQL parser and transpiler

Modin 0.34.0 - Make your pandas code run faster by changing one line of code.

pyOCD 0.37.0 - Cortex-M debugger for Python

Want More? Click Here

The Python Language Summit 2025

The Python Language Summit 2025 occurred on May 14th in Pittsburgh, Pennsylvania. Core developers and special guests from around the world gathered in one room for an entire day of presentations and discussions about the future of the Python programming language.

DjangoCon Europe 2025 Videos

PEP 685: Comparison of Extra Names for Optional Distribution Dependencies (Final)

PEP 751: A File Format to Record Python Dependencies for Installation Reproducibility (Final)

Articles

Optimizing Test Execution: Running live_server Tests Last with Pytest

It’s a good idea to run slow tests last; this way, you catch failed tests early and avoid putting unnecessary stress continuously on resources like databases and external services. In this article, Tim Kamanin shared how to achieve this by creating a custom marker and a pytest plugin tailored for Django application tests.

# Tim Kamanin

Navigating Imbalanced Datasets with Pandas and Scikit-learn

Imbalanced datasets are a common challenge in real-world machine learning. If not handled properly, models tend to favour the majority class, leading to biased results. In this article, Iván Palomares Carrascosa shared three effective strategies(Inverse Frequency, Undersampling, and Oversampling) to balance such data using Pandas and Scikit-learn.

# Iván Palomares Carrascosa

How Can You Structure Your Python Script?

Many people write Python scripts to automate tasks, but without proper structure, these scripts can quickly become hard to maintain. If you're planning to write a script, check out this Real Python article by Rohit Goswami. Your future self (and your teammates) will thank you. He explained how to organise your code using functions, constants, enums and data classes; improve usability with command-line arguments; logging; and even make scripts self-contained using PEP 723 to handle dependencies inline.

# Rohit Goswami

Beyond Model Stacking: The Architecture Principles That Make Multimodal AI Systems Work

I hope you remember in the Iron Man movie when Tony Stark flew around and JARVIS could understand what was in front of him? That kind of real-time visual understanding feels like science fiction, but Eric Chung wanted to bring it closer to reality. So he built VisionScout, a system combining YOLOv8 for object detection, CLIP for semantic reasoning, Places365 for scene classification and Llama 3.2 for language generation. He shared the full journey, along with insights and implementation details, in his article on Towards Data Science.

# Eric Chung

In this video, Vincent D. Warmerdam and Johnny demoed some of the powerful features of Marimo Notebook and walked through the amazing things you can build with it. It’s a great showcase of how Marimo stands out from traditional notebooks.

# Vincent D. Warmerdam & Johnny

Interesting Packages and Projects to Explore

Dishka - Cute DI framework with scopes and agreeable API

thepi.pe - Get clean data from tricky documents, powered by VLMs

Manim - Animation engine for explanatory math videos

Extractous - Fast and efficient unstructured data extraction. Extractous Python Binding

Bagels - Powerful terminal expense tracker.

Recent Noteworthy Package Releases

python-calamine 0.4.0 - Python binding for Rust's library for reading excel and odf file - calamine

SeleniumBase 4.40.0 - A complete web automation framework for end-to-end testing

pylance 0.31.0 - Python wrapper for Lance columnar format

PyAV 15.0.0 - Pythonic bindings for FFmpeg's libraries

PEFT 0.16.0 - Parameter-Efficient Fine-Tuning (PEFT)

CrewAI 0.140.0 - Cutting-edge framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks

statsig-python-core 0.6.0 - Statsig Python bindings for the Statsig Core SDK

haystack-experimental 0.11.0 - Experimental components and features for the Haystack LLM framework

wandb 0.21.0 - A CLI and library for interacting with the Weights & Biases API.

fastmcp 2.10.0 - The fast, Pythonic way to build MCP servers.

feast 0.50.0 - The Open Source Feature Store for AI/ML

sentence-transformers 5.0.0 - Embeddings, Retrieval, and Reranking

PaddlePaddle 3.1.0 - Parallel Distributed Deep Learning

pillow-heif 1.0.0 - Python interface for libheif library

bleak 1.0.0 - Bluetooth Low Energy platform Agnostic Klient

browser-use 0.4 - Make websites accessible for AI agents

PostHog 6.0.0 - Integrate PostHog into any python application

Want More? Click Here

A modern, opinionated, cookiecutter template for your next Python project

If you're starting a new project, it's smart to use a good template instead of building everything from scratch, as it can save a lot of time. There are plenty out there and if you're looking for a good one, check out the template recently created by Wyatt Ferguson.

# Wyatt Ferguson

Python 101 - An Intro to Working with INI files Using configparser

Did you know Python’s standard library includes configparser, which lets you work with Windows-style INI files? If not, check out this article by Mike Driscoll, where he walks through creating, editing and reading INI files with clear examples.

# Mike Driscoll

Text search is both a tricky and interesting topic. While tools like OpenSearch or Elasticsearch make it easier, building one yourself can teach you a lot. If you're up for that challenge, check out this tutorial video by Rahul Jha where he built a Wikipedia Search Engine from scratch using Gensim, TF-IDF and Flask.

# Rahul Jha

Python For Nonprofits to retrieve, analyze, visualize and share nonprofit data

In this repository, Kenneth Burchfiel walked through how to use Python to import, clean and reformat data, perform analysis using descriptive stats and linear regression, create charts and maps and share spreadsheets and interactive visualizations online. While it’s aimed at the nonprofit sector, it offers valuable lessons for anyone looking to strengthen their data presentation skills.

# Kenneth Burchfiel

Design Patterns You Should Unlearn in Python-Part2

There’s no silver bullet in life and the same goes for programming - even widely praised best practices like design patterns don’t apply the same way across languages. Python, for example, often offers simpler, more natural solutions than traditional patterns used in languages like C++ or Java. In this article, Cheng/Racey Chan covered two patterns - Flyweight and Prototype. If you resonate with his approach, don’t miss part 1 of his series as well.

# Cheng/Racey Chan

Interesting Packages and Projects to Explore

Shapely - Manipulation and analysis of geometric objects

Partial JSON Parser - Parse partial JSON generated by LLM

Mesa - Agent-based modeling (ABM) in Python

nanodjango - Run Django models and views from a single file, and convert it to a full project.

LiteLLM - Library to easily interface with LLM API providers

Recent Noteworthy Package Releases

Gymnasium 1.2.0 - A standard API for reinforcement learning and a diverse set of reference environments (formerly Gym)

LangGraph 0.5.0 - Building stateful, multi-actor applications with LLMs

Dagster 1.11.0 (core) / 0.27.0 (libraries) - An orchestration platform for the development, production, and observation of data assets.

aioboto3 15.0.0 - Async boto3 wrapper

lxml 6.0.0 - Powerful and Pythonic XML processing library combining libxml2/libxslt with the ElementTree API

transformers 4.53.0 - State-of-the-art Machine Learning for JAX, PyTorch and TensorFlow

mcp 1.10.0 - Model Context Protocol SDK

resolvelib 1.2.0 - Resolve abstract dependencies into concrete ones

chdb 3.4.0 - An in-process SQL OLAP Engine powered by ClickHouse

Diffusers 0.34.0 - State-of-the-art diffusion in PyTorch and JAX

junitparser 4.0.0 - Manipulates JUnit/xUnit Result XML files

Pybtex 0.25.0 - A BibTeX-compatible bibliography processor in Python

Instructor 1.9.0 - structured outputs for llm

Robyn 0.70.0 - A Super Fast Async Python Web Framework with a Rust runtime

Want More? Click Here

Registration for PyOhio 2025 is open!

Python 3.14.0 beta 3 is here!

This release, 3.14.0b3, is the third of four planned beta releases.

Articles

Building a Modern Python API with Azure Cosmos DB

Gwyneth Peña-Siguenza created a 5-part video series on building scalable Python APIs with FastAPI and Azure Cosmos DB. She covered key concepts like Pydantic models, FastAPI's dependency injection, async calls using azure.cosmos.aio, batch operations and centralised exception handling.

# Gwyneth Peña-Siguenza

How Reladiff Works

Pinpointing differences between two tables is very important for tasks like validating data migrations or spotting corruption. But when those tables live in different databases, it becomes tricky due to issues like network costs and different SQL dialects. In this article, Erez Shinnan shared how Reladiff tackles these challenges and its development journey.

# Erez Shinnan

Type hinting, type checking and data validation are often confusing words if someone uses them together, because they serve different purposes. In this video, Corey Schafer clearly explained the differences using a simple, practical example - it helps you understand what each term means, when to use them and how they work together in Python.

# Corey Schafer

t-strings are a new feature which is coming in Python 3.14. If you're wondering why we need t-strings when f-strings already exist, check out this video by Anthony Shaw. He explained how t-strings make it easier to create reusable templates and add custom logic or sanitisation, which makes string formatting more powerful and secure, with a clear and practical example.

# Anthony Shaw

Python is Removing GIL, Gradually

You've probably heard about the GIL vs no-GIL debate; it’s been happening for a while. If you’re not familiar with the full context, you should read this article by Yang Zhou. He explained what the GIL is, why it became a problem, what PEP 703 proposes, how to set up a no-GIL version of Python and what all of this means for Python developers. It's a great overview of a major topic in Python’s future.

# Yang Zhou

Interesting Packages and Projects to Explore

Click - Composable command line interface toolkit

pytest-xdist - pytest plugin for distributed testing, most importantly across multiple CPUs

DearPyGUI - A fast and powerful Graphical User Interface Toolkit for Python with minimal dependencies

Apache Tika - Python binding to the Apache Tika™ REST services allowing Tika to be called natively in the Python

pydantic-resolve - business model friendly data orchestration tool

Recent Noteworthy Package Releases

webdataset 1.0.2

Kaleido 1.0.0

Outlines 1.0.0

Valkey GLIDE 2.0.0

Streamlit 1.46.0

pydantic-graph, pydantic-evals, pydantic-ai-slim, pydantic-ai, fasta2a 0.3.0

oauthlib 0.3.0

labelbox 7.0.0

django-redis 6.0.0

ruff 0.12.0

opensearch-py 3.0.0

Multiple Azure client libraries bumped to major versions

bayesian-optimization 3.0.0

tcod 19.0.0

pyzmq 27.0.0

Want More? Click Here

Python 3.13.5 is now available

PyCon Africa CFP is open!

PyData Virginia 2025 Videos

Articles

Narwhals: Unified DataFrame Functions for pandas, Polars, and PySpark

Pandas used to be the go-to tool for working with data. But now, there are many other excellent options available, like Polars and PySpark. If different teams in your company use different data tools, Narwhals can help. It allows everyone to work with data in a consistent way, no matter which library they prefer. Marco Gorelli showed this with a helpful example in his article.

# Marco Gorelli

Writing a Text Editor in 7 Minutes using Textual

No Vim? No problem! Fredrik Sjöstrand got your back. You can build your own text editor with syntax highlighting using Textual and he showed how in his article.

# Fredrik Sjöstrand

Tiny Agents in Python: a MCP-powered agent in ~70 lines of code

The Hugging Face team has created tiny-agents, a new feature that lets their huggingface_hub software act as a Model Context Protocol (MCP) Client. This means that your AI models (LLMs) can now easily access and use tools from various MCP servers. In their recent article, they explained how to set up these tiny agents in Python to connect with MCP servers, which then gives your LLMs powerful new abilities to interact with the world and perform complex tasks.

# Hugging Face team

Marimo: A Reactive, Reproducible Notebook

Marimo notebooks are changing how we use notebooks because they are reactive. If you're interested in learning more, check out the Real Python article by Ian Eyre. He covered a lot about Marimo notebooks, including How to create a Marimo notebook, Marimo's reactivity, Adding interactivity to notebooks, How to share your notebooks.

# Ian Eyre

Avoid indexes in Python

While indexes are useful, relying on them too much can be like Maslow's hammer – if all you have is a hammer, everything looks like a nail. In other words, you might try to solve every problem with the same tool, even when there are better ways. Trey Hunner has shown some fantastic alternative methods for common tasks like checking if something is in a list, getting specific items from tuples and looping through multiple lists, all without constantly needing to use indexes.

# Trey Hunner

Interesting Packages and Projects to Explore

tqdm - ⚡ A Fast, Extensible Progress Bar for Python and CLI

iPOPO - A service-oriented component model framework

nbconvert - Converting Jupyter Notebooks (.ipynb files) to other formats.

Arrest - Arrest is a wrapper of pydantic and httpx to make your REST api calls type-safe and structured

Ovld - Overloading Python functions

Recent Noteworthy Package Releases

NumPy 2.3.0

google-adk 1.3.0

pip-system-certs 5.0

django-multiselectfield 1.0.0

shap 0.48.0

django-waffle 5.0.0

schemathesis 4.0.0

Want More? Click Here

Monkey-Patched PyPI Packages Use Transitive Dependencies to Steal Solana Private Keys

Malicious PyPI package semantic-types steals Solana private keys via transitive dependency installs using monkey patching and blockchain exfiltration.

Python 3.13.4, 3.12.11, 3.11.13, 3.10.18 and 3.9.23 are now available

Registration is now open for Python for Good

Python Discord Event: Project Show-and-Tell

This will be an opportunity for community members to present their Python projects.

Articles

Decorators and Functional programming

In this article, Racey Chan explained key functional programming concepts: function composition, currying, closures, etc, using Python decorators through practical examples.

# Racey Chan

Python RAG API Tutorial with LangChain & FastAPI – Complete Guide

In this article, Vitalii Honchar explained how to build AI-powered apps that can chat with uploaded PDF files. He showed how to implement Retrieval Augmented Generation (RAG) using FastAPI for the API and LangChain to interact with OpenAI.

# Vitalii Honchar

Nested Loops in Python

Nested loops are often discouraged, but they make perfect sense in scenarios like handling multidimensional data or generating patterns. In this article, Amanda Adoyi explained what nested loops are, shared practical examples, highlighted common pitfalls and offered tips on how to optimise them.

# Amanda Adoyi

Loading Pydantic models from JSON without running out of memory

When working with large JSON files and loading them into Pydantic, memory usage can become a concern. Instead of using model_validate_json, it's better to use parsers like ijson to handle data efficiently. In this article, Itamar Turner-Trauring explained how to do that and also shared another memory-saving approach using dataclasses with __slots__ for better performance.

# Itamar Turner-Trauring

Pyrefly vs. ty: Comparing Python’s Two New Rust-Based Type Checkers

Last month, two new Rust-based Python type checkers pyrefly and ty were released. Both of them are in the alpha stage. While they share some similarities, they differ significantly in design and features. In this article, Edward Li dove deep into both tools, highlighted their differences and what makes each one unique.

# Edward Li

Interesting Packages and Projects to Explore

asyncpg - A fast PostgreSQL Database Client Library for Python/asyncio.

PAR LLAMA - Terminal UI for Ollama and other LLM providers

Fabric - High level SSH command execution

Dataclasses JSON - Easily serialize dataclasses to and from JSON.

Copier - Library and command-line utility for rendering projects templates.

Recent Noteworthy Package Releases

scikit-learn 1.7.0

delta-spark 4.0.0

compressed-tensors v0.10.0

langfuse 3.0.0

crewai 0.126.0

LanceDB v0.23.0

opentelemetry-sdk 1.34.0

wandb v0.20.0

deepspeed v0.17.0

py7zr 1.0.0

ffmpy 0.6.0

Want More? Click Here

Python 3.14.0 beta 2 is here!

This release, 3.14.0b2, is the first of four planned beta releases.

PyCon NL CFP is Open

You have the chance until August 1, 2025, to submit. If you are looking to volunteer, then visit this link.

Articles

Namespaces in Python

Variables and functions have names and we use those names to refer to them, but how does Python actually keep track of them? The answer lies in namespaces. Leodanis Pozo Ramos explained this concept clearly in this Real Python article. He covered what a namespace is in Python, the concept of scope, the LEGB rule for resolving names, how globals() and locals() work, how to modify variables from different scopes using global and nonlocal, etc.

# Leodanis Pozo Ramos

Arjan Egges shared 10 common Python anti-patterns that often lead to unnecessarily complex code and explained how to avoid them with better alternatives. Some key suggestions: include using modules instead of classes with only static methods, avoiding unnecessary overrides of dunder methods and relying on Python’s built-in tools and libraries instead of reinventing solutions.

# Arjan Egges

Python Tooling at Scale: LlamaIndex’s Monorepo Overhaul

The LlamaIndex ecosystem has over 650 Python packages in a monorepo, making dependency management and publishing a challenge. To solve this, the team built LlamaDev and used uv it for improved performance. In this post, Massimiliano Pippi explained the issues they faced, what they tried and why they ultimately created LlamaDev.

# Massimiliano Pippi

LangGraph is a powerful library for creating advanced conversational AI workflows. If you're interested in learning how to use it and want to build your own AI agent, check out the freeCodeCamp course by Vaibhav Mehra. It's a great starting point to explore LangGraph in action.

# Vaibhav Mehra

From Numbers to Narrative: How I Use Data Science to Decode Market Trends

If you’re looking to start building prediction models, stock price prediction is a great choice; it can be as simple or as complex as you want. A good starting point is this article by Altamash Ansari, where he shared how to build a basic model that predicts stock prices using historical data.

# Altamash Ansari

Interesting Packages and Projects to Explore

pyTermTk - Python Terminal Toolkit - a Spiced Up Cross Compatible TUI Library 🌶️

structlog - Structured Logging for Python

Opik - Debug, evaluate, and monitor your LLM applications, RAG systems, and agentic workflows with comprehensive tracing, automated evaluations, and production-ready dashboards

rpy2 - Python interface to the R language (embedded R)

Fused - A Python library to code, scale, and ship geospatial workflows of any size.

Recent Noteworthy Package Releases

Deltalake 1.0.0

DeepEval v3.0

pytest-asyncio 1.0.0

Curlify v3.0.0

cachetools v6.0.0

Apache Spark 4.0.0

Want More? Click Here

Python: The Documentary [OFFICIAL TRAILER]

PyCA cryptography 45.0 released

PyData Amsterdam CALL FOR PROPOSALS

PyData/Sparse 0.17.0 released

PyCon US 2025 Videos Playlist

web3.py v7.12.0 released

Articles

Making PyPI's test suite 81% faster

The Trail of Bits team managed to cut down PyPI’s test suite runtime from 163 seconds to just 30 - a remarkable 81% improvement. In their write-up, they detailed the steps that made it possible: using pytest-xdist for parallel test execution, leveraging Python 3.12’s sys.monitoring, configuring testpaths, removing unnecessary imports, etc.

# Trail of Bits

Interactive reStructuredText Tutorial

Looking to learn or practice reStructuredText? This interactive tutorial site is a great place to start. It covers everything from basic syntax and inline markup to advanced topics like directives, roles, tables and figures - all in a hands-on playground format.

# yakimka

SQLAlchemy 2.0: The Most Powerful ORM for Python Yet

SQLAlchemy 2.0 has been around for a while and it brings some major changes compared to 1.x, like replacing query with select for fetching data. In this article, the Leapcell team gave a thorough walkthrough of how to use SQLAlchemy from setting up a database connection to performing batch inserts.

# Leapcell

Django Security Best Practices: A Comprehensive Guide for Software Engineers

A vulnerable web app can lead to data loss, loss of user trust and even legal trouble. That’s why securing it is crucial. In this article, Ahmad Sadeddin shared Django security best practices like enforcing HTTPS, enabling Django’s built-in security features, adding rate limiting and more.

# Ahmad Sadeddin

The first year of free-threaded Python

Nathan Goldbaum shared the story of the first year of efforts on Python’s free-threaded build about how the Quansight team played a key role in making it usable with real production workflows, even those involving complex dependency stacks.

# Nathan Goldbaum

Interesting Packages and Projects to Explore

nglview - IPython widget to interactively view molecular structures and trajectories.

AG2 (formerly AutoGen) - A programming framework for agentic AI

nbdev - Create delightful software with Jupyter Notebooks

tyro - CLI interfaces & config objects, from types

Django Ninja - A web framework for building APIs with Django and Python 3.6+ type hints.

About Upcoming Python Events

PyCon Italia 2025
May 28 – 31, 2025

PyLadies Amsterdam: Dapr in Action: From Core Concepts to AI Agents
May 27, 2025

By The Same Team Behind FastAPI - FastAPI Cloud — You code. We Cloud.

FastAPI creator Sebastián Ramírez announced FastAPI Cloud, a service to streamline deployment of FastAPI applications.

Announcing Python Software Foundation Fellow Members for Q1 2025!

Django security releases issued: 5.2.1, 5.1.9 and 4.2.21

🇰🇷 🎙️ PyCon KOREA CFP OPEN

Introducing Pyrefly: A new type checker and IDE experience for Python

Facebook team announced an alpha version of Pyrefly, an open-source Python type checker and IDE extension crafted in Rust.

Articles

How to Add Blazing Fast Search to Your Django Site with Meilisearch

A strong search feature improves user experience, especially on e-commerce sites where customers might only remember part of a product name or its use. Smart autocomplete and relevant suggestions not only save time but also boost sales. In this blog post, Lacey Henschel shared how she built just that for a client using Meilisearch with Django, walked through the setup and implementation with a clear example.

# Lacey Henschel

Python project management and packaging pep 751 update and some of the remaining issues of packaging

PEP 751 has been accepted. A big step toward making Python dependency installation more reproducible. If you want to deeply understand why this matters, what problems it solves, and how it actually works, this article is a great place to start.

# Reinforced Knowledge

Asyncio Demystified: Rebuilding it From Scratch One Yield at a Time

If you really want to understand a concept deeply, building a simple version of it can help a lot. Jb Rocher took this approach to learn how asyncio works. In his article, he walked through creating a basic version of asyncio, showing how generators and futures are its core building blocks. It’s a great read if you want to demystify asynchronous programming in Python.

# Jb Rocher

Module enum overview

Enums can make your code more readable, safer and easier to maintain. If you want a solid overview of Python’s enum module, check out this article by Rodrigo. He covered everything from basic Enum usage to auto, StrEnum, Flags and more. It’s a great way to learn the full range of what Python enums offer.

# Rodrigo

How I Created an MCP Server for PostgreSQL to Power AI Agents — Components, Architecture & Real Testing

Model Context Protocol (MCP) helps AI models connect with real-world data, tools and services. If you're thinking about building your own MCP server to read files, access databases or interact with local tools, check out this article by Gaurav Kumar. He showed how to create a simple MCP server for PostgreSQL.

# Gaurav Kumar

Interesting Packages and Projects to Explore

ASE - Atomic Simulation Environment: A Python library for working with atoms

sktime - A unified framework for machine learning with time series

SnakeViz - A web-based viewer for Python profiler output

Shiny - A web development framework for Python.

PyAudio - PyAudio provides Python bindings for PortAudio, the cross-platform audio I/O library.

About Upcoming Python Events

PyData Delhi x PyDelhi Meetup
May 17, 2025

Django Meetup Vol. 66 / Decoupling the Monolith
May 20, 2025

Inland Empire Python Users Group monthly meeting
May 20, 2025

Python 3.14.0 beta 1 is here!

Python 3.14 is still in development. This release, 3.14.0b1, is the first of four planned beta releases.

Ty: An extremely fast Python type checker and language server, written in Rust

Released a couple of days ago by Astral, it has already hot topic in many places and got a lot of GitHub Stars too.

Textualize, the company, will be wrapping up in the next few weeks

It’s sad news, but unfortunately, we need to accept.

Articles

10 Free Data Science Books For 2025

Looking to level up your data science skills? Cornellius Yudha Wijaya has shared an excellent list of free data science books in this article - perfect for expanding your knowledge without spending a dime.

# Cornellius Yudha Wijaya

Python is an interpreted language with a compiler

We often hear Python is an interpreted language and Java is a compiled one. So, does that mean Python doesn’t use a compiler and Java doesn’t use an interpreter? Not really. In this article, Nicole Tietz-Sokolskaya explained clearly. It might not answer everything, but it will help you better understand how programming languages really work or give you new questions to explore.

# Nicole Tietz-Sokolskaya

Debugging Python f-string errors

We all love f-strings in Python. You might think f"{x}" is just the same as str(x) - and that's mostly true, but not completely. Behind the scenes, f"{x}" actually calls format(x, "") or type(x).__format__(x, ""). Brandon Chinn ran into a TypeError while using an f-string, which led him to investigate what was really going on. He shared what he discovered in this interesting blog post.

# Brandon Chinn

14 Advanced Python Features

When someone asks about advanced Python features, most people mention generators, comprehensions, OOP or asyncio. But Python has a lot more to offer. In this article, Edward Li shared 14 unique, underrated advanced features - like typing overloads, protocols and more - each explained with clear examples.

# Edward Li

A practical guide to building agents

Thinking of creating your first agent? Check out this guide by OpenAI - it’s packed with practical tips and best practices. It also covers frameworks for designing agent logic and orchestration and how to ensure your agents run safely, predictably and effectively.

# OpenAI

Interesting Packages and Projects to Explore

Skulpt - It is a Javascript implementation of the Python programming language

Flower - Real-time monitor and web admin for Celery distributed task queue

DBOS - Ultra-lightweight durable execution in Python

Distlib - A low-level library which implements some Python packaging standards (PEPs) and which could be used by third-party packaging tools to achieve interoperability.

TkForge - Drag & drop in Figma to create a Python GUI with ease

About Upcoming Python Events

PyCon US 2025
May 14 – 22, 2025

Leipzig Python User Group Meeting
May 13, 2025

PythonCDMX 🐍 MAY 2025 - Publicando un paquete en PyPI
May 13, 2025

PyCologne Discussions
May 14, 2025

Python Ireland May 2025 Meetup
May 14, 2025

PiterPy 2025
May 16 – 17, 2025

PyGrunn 2025
May 16, 2025

PyCaxias 2025
May 17, 2025

BangPypers - Bangalore Python Users Group
May 17, 2025

Django Girls Caxias do Sul 2025
May 18, 2025

PyLadies Dublin May Meetup @ Pinterest
May 20, 2025

Santa Cruz Python Meetup: Vibe Coding: why, when, how
May 22, 2025

PEP 787: Safer Subprocess Usage Using t-strings (Postponed to 3.15)

PyData Global 2024 videos are out

PEP 790 – Python 3.15 Release Schedule

Articles

What's new in pip 25.1 - Dependency groups!

Pip 25.1 is out with some exciting updates! This release introduced a new install progress bar, experimental support for lockfile generation with pip lock (PEP 751) and support for Dependency Groups (PEP 735). Richard shared these features and more in his latest blog post - definitely worth a read if you want to stay up to date.

# Richard

Bootstrapping Python projects with copier

If you frequently create new Python projects, you’ve probably used cookiecutter templates. But if you're looking for an alternative with more flexibility, copier package is a great option. In this article, Tucker Beck showed how to build and use project templates with copier.

# Tucker Beck

We hacked Google’s A.I Gemini and leaked its source code (at least some part)

Roni "Lupin" Carta shared their hacking journey targeting Google’s latest AI, Gemini, as part of the bugSWAT program. In the write-up, he detailed their approach, the techniques they used and how they ultimately discovered a vulnerability.

# Roni "Lupin" Carta

How to run Python in production

When preparing to run your Python application in production, taking a few key precautions can save you from major issues later, like preventing secret leaks, ensuring legal compliance and more. In this article, Ashish Bhatia shared some of the important steps to consider before going live.

# Ashish Bhatia

Getting Started with Streamlit

Streamlit has made it easy for data scientists to turn their work into shareable web apps. If you're just getting started, check out this beginner-friendly tutorial by Martin Fitzpatrick. He walked through adding titles, headings, text, buttons, dropdowns, sliders and a sidebar. Gave a solid idea to build simple and interactive apps.

# Martin Fitzpatrick

Interesting Packages and Projects to Explore

pyftpdlib - Extremely fast and scalable Python FTP server library

Param - Make your Python code clearer and more reliable by declaring Parameters

Edifice - Declarative GUI framework for Python and Qt

pytest-benchmark - A pytest fixture for benchmarking code. It will group the tests into rounds that are calibrated to the chosen timer.

Manim - A community-maintained Python framework for creating mathematical animations.

About Upcoming Python Events

PyTexas Monthly Meetup
May 7, 2025

Python Atlanta Meetup
May 8, 2025

BookQuest - Gamifying Libraries for Kids with Python workshop
May 10, 2025

Bamenda - Creating Python Communities and outreach
May 10, 2025

PEP 750 – Template Strings accepted

Wikipedia is giving AI developers its data to fend off bot scrapers

If you're planning to crawl Wikipedia for AI training, you might want to skip the hassle - Wikipedia has released dataset on Kaggle that's specifically optimised for training AI models.

Apache Airflow 3.0 is out!

Articles

How to Optimize your Python Program for Slowness

We all aim to make our programs faster, but have you ever tried doing the opposite? I mean, intentionally making them slower. Understanding how to write painfully slow code can actually help you recognise performance pitfalls. In this article, Carl Kadie demonstrated how a simple nested loop can create a program that runs longer than the lifetime of the universe. He dove into concepts like tetration (yes, it goes beyond exponentiation), 5-State Turing Machines and how Python’s int type can drastically affect performance.

# Carl Kadie

This course by Vuk Rosić gives a complete deep dive into DeepSeek V3, a state-of-the-art deep learning model. He covered theoretical explanations with step-by-step coding instructions to make things easier to understand and implement the model from scratch.

# Vuk Rosić

An Introduction to Testing in Python Flask

Writing tests is one of the best ways to ensure code quality, streamline development and boost confidence when making changes. If you're looking to learn how to write tests for your Flask application, check out this article by Federico Trotta. He walked through everything from writing basic tests to mocking external services, It’s a great resource for levelling up your Flask testing skills.

# Federico Trotta

Mutable default arguments

Default values in function arguments are super handy, but using mutable objects (like lists or dictionaries) as defaults can cause unexpected behaviour, because they’re shared across calls. Even beyond mutables, shared default values can lead to bugs. If you still want to use them safely, check out this article by Trey Hunner. He explained them with clear examples and offered practical solutions.

# Trey Hunner

Building Transformers from Scratch

Transformers are at the core of all modern LLMs. Understanding how they work can deepen your grasp of these powerful models. If you're curious to learn more, check out this article by Stephen Diehl. He walked through building transformers from scratch and explained the concepts in a hands-on, approachable way.

# Stephen Diehl

Interesting Packages and Projects to Explore

Python-Blosc2 - A fast & compressed ndarray library with a flexible compute engine.

PipeFunc - A Python library for defining, managing, and executing function pipelines.

Narwhals - Extremely lightweight compatibility layer between dataframe libraries

pytest-testmon - selects tests affected by changed files and methods

Patito - A dataframe modelling library built on top of polars and pydantic.

About Upcoming Python Events

PyLadies Amsterdam: Open source, open pipelines: Data ingestion with modern data stack
April 29, 2025

PyCamp España 2025
May 1 – 4, 2025

🐍 Canberra Python Meetup - Use Python as your shell with Xonsh by Leo Broska
May 1, 2025

Michigan Python May Meeting
May 1, 2025

NVIDIA Finally Adds Native Python Support to CUDA

For years, NVIDIA’s CUDA toolkit lacked native Python support, but that’s finally changed.

Pydantic v2.11 Released!

This release brought significant performance boosts and powerful new features that can make a real difference in your projects. It might be old news, but it's definitely worth knowing.

Articles

Django: what’s new in 5.2

Django 5.2 was released on April 2nd, and it brought some exciting new features. In this article, Adam Johnson shared several highlights with examples - like automatic model imports in the shell, support for composite primary keys, a simplified way to override BoundField, etc.

# Adam Johnson

The final alpha release for Python 3.14 dropped on April 8th, which means no new features will be added from here on out. The official release is set for October - yep, Python 3.14 (pi 😉). If you're curious about what's coming—like PEP 765, PEP 758, PEP 649, performance boosts and more—check out this great breakdown video by Denis Gruzdev.

# Denis Gruzdev

AI agents are making tasks like data scraping much easier. If you're exploring options, the Browser Use package is a solid one to check out. In this video, Tim Ruscica (aka Tech With Tim) showed how to use it and how it simplifies the scraping process. This tool/package is not perfect, but it is definitely a great starting point for experimentation.

# Tim Ruscica (aka Tech With Tim)

Outlier Detection with Python

Outliers can reveal everything from fraud to fascinating patterns in your data. In this episode of Talk Python to Me, Brett Kennedy dove into outlier detection using Python world - covered many things related to it. A great listen if you're curious about spotting the unusual in your datasets.

# Brett Kennedy

Optimizing Python: Understanding Generator Mechanics, Expressions, And Efficiency

Python generators provide a clean and memory-efficient way to handle iteration. They’re especially useful for tasks like processing large datasets, handling infinite sequences or reading big files without loading everything into memory. If you want to dive deeper into how and when to use them, check out this great article by Josh Engroff.

# Josh Engroff

Interesting Packages and Projects to Explore

py-spy - Sampling profiler for Python programs

Robyn - A Super Fast Async Python Web Framework with a Rust runtime.

PyFunctional - Package for creating data pipelines with chain functional programming

airspeed velocity (asv) - A simple Python history benchmarking tool

Mininterface - A minimal interface to a Python application (GUI, TUI, CLI, web)

About Upcoming Python Events

DjangoCon Europe 2025
April 23 – 27, 2025

PyCon DE & PyData 2025
April 23 – 25, 2025

PyCon Lithuania 2025
April 23 – 25, 2025

PyKla monthly meetup
April 19, 2025