Python Releases

Python language news, releases and ecosystem updates. New features, breaking changes, security advisories and deprecations - each summarised in plain English and updated continuously.

Tracking 276 Python releases · Updated

  • Python Insider blogengineer ·

    Python Documentation Now Available in German

    The Python documentation is now available in German, significantly enhancing accessibility for German-speaking users worldwide. This milestone was achieved through dedicated community contributions, making Python more inclusive. The project actively seeks more volunteers to help translate and maintain documentation in various other languages. Interested contributors can find details in the Python Developer's Guide and the Translation dashboard.

    announcement
  • Pydantic AI Releases sdkaiazureengineerpython-agentic ·

    pydantic-ai v2.51.0 Adds OpenAI GPT-Live Support and Refines AI Model Compatibility

    pydantic-ai v2.51.0 introduces support for OpenAI GPT-Live through `OpenAILiveModel` and exposes `context_window_used` on realtime sessions, providing better visibility into token usage. The update also includes compatibility adjustments for Gemini 3.1 Flash Live and 3.8 Live, specifically rejecting dated IDs and raising `RealtimeError` for close codes. Additionally, it improves error handling by raising `UserError` for forced tool calls in OpenAI, Azure OpenAI, and xAI models, alongside several bug fixes.

    breaking feature patch
  • Google ADK (Python) Releases sdkaigcpdeprecationengineermediapython-agentic ·

    Google ADK for Python v2.10.0 Boosts Skill Lifecycles, DB Integrations, and Eval Metrics

    Google ADK for Python v2.10.0 introduces advanced skill lifecycle management, expanded database and model integration support, and richer efficiency metrics for agent evaluation. This release provides developers with dynamic control over agent resource usage, seamless vector and hybrid search capabilities via a new MongoDB toolset, and detailed insights into execution costs and latency. It enhances integration with OpenAI reasoning models and includes several behavior changes for BigQuery protected mode. Additionally, legacy live-audio modules are now deprecated, affecting engineers building sophisticated AI agents.

    breaking deprecation feature patch
  • Python PEPs infraengineer ·

    PEP 848 Proposes Generational Incremental Garbage Collection for CPython

    PEP 848 introduces a proposal for a new generational and incremental cyclic garbage collector in CPython, targeting Python 3.16. This collector aims to significantly reduce GC pause times, potentially by up to 100-fold in large-heap programs, and improve overall performance by 3-5% on benchmarks. It achieves this by focusing collections on parts of the heap where garbage is more likely to accumulate, addressing inefficiencies of the current collector. The existing generational collector will remain an available option for developers.

    announcement
  • Pydantic AI Releases sdkaigaengineerpython-agentic ·

    Pydantic-AI v2.50.0 Introduces DecisionModel and Gemini Realtime Support

    Pydantic-AI v2.50.0 introduces the new `DecisionModel` base, designed to streamline AI model routing by name with labeled routes and enhanced context for improved decision-making workflows. This release also adds real-time support for `gemini-3.8-live` and `gemini-3.8-live-extended-thinking` models, expanding integration options and capabilities. Several compatibility notes are included, particularly affecting `DecisionModel` usage and `ModelSelectionContext` messages, which may require developer adjustments. The update also delivers numerous bug fixes, including corrected pricing for Anthropic and real-time image inputs, along with improvements for streaming audio and error handling.

    breaking feature patch
  • uv Releases sdkpreviewengineer ·

    uv 0.12.19 Released with PyPy/GraalPy Updates and New Preview Features

    uv version 0.12.19 has been released, expanding Python compatibility with support for new PyPy 3.11.16/3.12.14 and GraalPy 3.13.0 versions. The update introduces preview features for lazy imports in build-backend hooks on CPython 3.15+ and optimized lockfile freshness checks. This release, targeting Python developers, also includes several bug fixes, enhancements to `uv publish` errors, and a restore to the Rust API's `FlatDistributions` export.

    feature patch
  • Django Weblog blogengineereducationpython-web ·

    Ken Whitesell Named DSF Member of the Month for September 2026

    The Django Software Foundation (DSF) has recognized Ken Whitesell as its Member of the Month for September 2026. Whitesell, a long-time contributor to the Django community, is celebrated for his extensive volunteer work, including over a decade with DjangoCon US and consistent support in the Django forum. The feature highlights his journey as a retired software developer, his passion for boardgaming, and his commitment to nurturing the community. This announcement celebrates his significant impact and the motivation behind his ongoing contributions.

    announcement
  • Pydantic AI Releases sdkaigaengineerpython-agentic ·

    Pydantic AI v2.49.0 Enhances LLM Integrations and Type-Safe Model Handling

    Pydantic AI v2.49.0 introduces several new features, including a GitHub Copilot OAuth flow for device authorization and improved `TypeSafeModel` capabilities for handling boolean fields and `None` values. This release also provides fixes for a range of bugs across `OpenAIChatModel`, AWS Bedrock Converse, Google Vertex AI, and `GraphBuilder` functionality. These updates aim to provide greater flexibility in model definition, enhance integration with various AI services, and improve the reliability of AI agent development. Developers leveraging Pydantic AI for LLM-powered applications will benefit from these advancements and stability improvements.

    feature patch
  • LangGraph Releases sdkaiengineerpython-agentic ·

    LangChain LangGraph CLI Updates to Version 0.4.32.dev0

    This update announces version 0.4.32.dev0 of the LangGraph command-line interface. LangGraph is a framework built on LangChain for building robust, stateful multi-actor applications with LLMs. While specific changes are not detailed in this announcement, the release indicates ongoing development and maintenance for the tool.

    patch
  • LangGraph Releases sdkaigaengineerpython-agentic ·

    LangGraph CLI 0.4.32 Enhances Deployments, Agent Flags, and Addresses Vulnerabilities

    LangGraph CLI version 0.4.32 introduces several new features for self-hosted deployments, including listener placement and an `--image-uri` flag. It also refines agent flag management for clearer configurations and environment defaults. This release incorporates dependency updates, notably addressing AnyIO vulnerabilities, and fixes issues related to missing deployment configuration clarity. These enhancements improve the flexibility, security, and developer experience for deploying LangGraph applications.

    security feature patch
  • Python Insider bloggaengineer ·

    Python Documentation Now Available in Persian

    Python's official documentation is now available in Persian (Farsi), expanding its accessibility to a global audience of developers. This significant milestone was achieved through dedicated community contributions, making Python more inclusive for non-English speakers. The project relies on volunteer efforts, and additional help is welcomed to maintain and update this and other language translations.

    announcement
  • Pydantic AI Releases sdkaigaengineerpython-agentic ·

    Pydantic-AI v2.48.0 adds support for new OpenAI and Claude models

    Pydantic-AI v2.48.0 introduces support for recent OpenAI audio models, including `gpt-6-sol` and `gpt-6-luna`, and Claude Opus 5.5. This update expands the library's compatibility with advanced generative AI models, allowing developers to leverage the latest capabilities. It also includes bug fixes for preserving `UIMessage.id` and `AG-UI Message.id` across message loading and dumping operations. This release primarily affects developers using Pydantic-AI for integrating with large language models and OpenAI's audio features.

    feature patch
  • uv Releases sdkinfrapreviewengineer ·

    uv 0.12.18: Security Fix for Windows Path Traversal, Pip Enhancements

    uv version 0.12.18 has been released, including a security fix for a path traversal vulnerability (GHSA-2cv4-cqwr-gwf7) affecting Windows during wheel installation. This update introduces `--output-format json` for `uv pip install` and `uv pip sync`, along with a new `--check` flag to report planned changes without modifying the environment. The release also delivers general enhancements, performance improvements, and bug fixes across the tool. Users on Windows are advised to update to mitigate the path traversal risk.

    security feature patch
  • Django Weblog blogengineerpython-web ·

    Django Project Approves New Technical Governance Model

    The Django Project's Steering Council and DSF Board have officially approved DEP 19, implementing a new technical governance model. This update simplifies the existing governance structure and introduces revised eligibility criteria for the Steering Council. The changes aim to make participation more approachable and encourage a wider range of community members to stand for election. Future steps involve updating documentation across Django repositories before the DEP reaches its final state.

    announcement
  • Python PEPs engineer ·

    PEP 849 Draft Proposes More Expressive Type Expressions in Python 3.16

    PEP 849 is a draft proposal to enhance Python's type annotation system, allowing any Python expression to be used in type annotations by returning the annotation's Abstract Syntax Tree (AST) and namespaces for runtime evaluation. This aims to overcome current limitations where many valid expressions cannot be introspected at runtime, restricting type expressiveness. If accepted, the PEP would pave the way for more intuitive type syntax like `1 | 2 | 3` for literal unions and conditional types, affecting Python developers who rely on advanced type checking and runtime introspection. It is currently in draft status for Python 3.16.

    announcement feature
  • Pydantic AI Releases sdkaigaengineerpython-agentic ·

    Pydantic-AI v2.47.0 Improves Type Safety and Error Handling for Model Outputs

    Pydantic-AI v2.47.0 has been released, introducing several enhancements and critical bug fixes. This update includes compatibility changes for output route naming and user prompt content validation, along with new capabilities for TypeSafeModel to answer more field types. Developers using Pydantic-AI will benefit from improved robustness in handling model outputs, better error messaging for invalid configurations, and more flexible routing options, ensuring greater stability in AI application development.

    breaking feature patch
  • Python PEPs engineer ·

    PEP 824 Proposes None-Coalescing Operators `??` and `??=` for Python 3.16

    A new Python Enhancement Proposal (PEP 824) outlines the addition of two None-coalescing operators, `??` and `??=`, to the language. These operators aim to simplify common patterns for handling `None` values, offering a more concise and predictable way to provide fallback values or conditionally assign them. Proposed as a draft for Python 3.16, the operators reduce boilerplate code and mitigate issues like unintended side effects from repeated expressions when checking for `None`. This initiative addresses long-standing community discussions for a dedicated null-coalescing mechanism, similar to those found in other modern languages.

    announcement feature
  • Python PEPs engineer ·

    PEP 823 Proposes None-Aware Access Operators for Python 3.16

    This Python Enhancement Proposal (PEP 823) introduces two new operators, `?.` for attribute access and `?[]` for indexing, designed to simplify handling of `None` values. These "None-aware" operators would evaluate the left-hand side, and only proceed if the result is not `None`, preventing `AttributeError` or `TypeError`. This aims to reduce boilerplate `is not None` checks and improve code readability, particularly for nested optional attributes or structured data traversal. The proposal is currently a draft targeting Python 3.16.

    security feature patch
  • LangGraph Releases sdkaigaengineerpython-agentic ·

    LangGraph 1.2.12 improves subgraph detection and adds interrupt response schema

    LangGraph version 1.2.12 introduces a new `response_schema` parameter to the `interrupt()` method, providing developers with more explicit control over interrupt handling in agent workflows. This release also includes a significant fix that enhances subgraph detection by analyzing bytecode instead of source code, improving accuracy and stability. Additionally, it addresses an issue with undeclared V3 stream projection types and updates several dependencies. These changes benefit engineers building and maintaining sophisticated agentic systems with LangGraph.

    feature patch
  • LangGraph Releases sdkaigaengineerpython-agentic ·

    LangGraph SDK 0.4.5 Adds `response_schema` to `interrupt()` Method

    LangGraph Python SDK version 0.4.5 has been released, introducing a `response_schema` parameter to the `interrupt()` method. This enhancement allows developers to define the expected schema for responses when interrupting graph execution, improving the robustness and predictability of AI agent interactions. The update also includes several dependency bumps, notably for `anyio` and `websockets`, contributing to overall stability and security. This release primarily affects engineers building AI agent applications that rely on structured responses during interactive graph execution.

    feature patch

About Python release tracking on ReleaseBytes

Python point releases, security fixes and end-of-life dates matter to almost every stack — including indirectly, through what Databricks runtimes and cloud services pin. ReleaseBytes tracks CPython releases and ecosystem news, and its EOL tracker shows exactly when each Python version stops receiving security fixes.

Frequently asked questions

How often are Python release notes updated on ReleaseBytes? ›

Continuously. ReleaseBytes monitors the official Python release channels around the clock and publishes a plain-English summary of each announcement shortly after it lands.

What kinds of Python changes does ReleaseBytes track? ›

New features, enhancements, bug fixes, security advisories, breaking changes, deprecations and end-of-life announcements. Every item is tagged by type so you can filter to just the changes that need action.

How can I get alerts for new Python releases? ›

Set up a free email or Slack alert filtered to Python, subscribe to the weekly digest, or follow the RSS feed. Teams can also install the ReleaseBytes GitHub App or connect via MCP.

Where does the Python release data come from? ›

From the official sources: Python language news, releases and ecosystem updates. Every item links back to the original vendor announcement.