Anthropic Releases
Anthropic news, Claude releases and SDK updates. New features, breaking changes, security advisories and deprecations - each summarised in plain English and updated continuously.
Tracking 218 Anthropic releases · Updated
- Anthropic News aipreviewengineer ·
Anthropic Beta: Claude Usage Reflection Dashboard
Anthropic has launched a beta feature allowing users to track and analyze their interaction patterns with Claude, aiming to improve AI integration into daily life. This reflection dashboard visualizes usage across key topics and task types, offering insights and interactive prompts to refine AI collaboration. It is available to Free, Pro, and Max users with Memory enabled, with specific privacy measures for sensitive conversations.
feature announcement - Claude Code Releases sdkaipreviewengineer ·
Claude Code v2.1.205: Bug fixes and usability improvements
This release of Claude Code, version 2.1.205, addresses numerous bugs and introduces several usability enhancements for its AI coding assistant. Key fixes include preventing silent data loss during agent interactions, resolving issues with file handling on Windows, and improving the reliability of background agent operations. The update also enhances the agent view with clearer status indicators and adds a diagnostic tool. These changes benefit developers and architects using Claude Code for coding assistance, aiming for a more stable and informative user experience.
patch - Claude Code Releases sdkaiengineermedia ·
Claude-Code v2.1.204: Fixes hook event streaming in headless sessions
Version 2.1.204 of Claude-Code resolves an issue where hook events failed to stream during SessionStart in headless sessions. This prevented remote workers from being prematurely idled. The fix ensures proper event propagation for headless operation.
patch - Claude Release Notes aipreviewengineer ·
Claude Cowork available on web and mobile
Claude Cowork is expanding to web and mobile platforms, complementing its existing desktop availability. This feature allows sessions and files to be saved to a user's Claude account, enabling access across devices and ensuring scheduled tasks run even when a device is offline. The rollout begins with the Max plan over the coming weeks, and all chat and Cowork features are now consolidated in a single location.
feature announcement - Claude Code Releases sdkaipreviewengineer ·
Claude Code v2.1.203: Session Recovery and UI Improvements
This update to Claude Code addresses numerous bugs related to background session stability and responsiveness, particularly on macOS and Windows. It improves the user experience by adding warnings for expiring logins, clearer visibility of active modes, and better recovery from session staleness. The release affects developers using Claude Code for coding assistance, especially those working with background agents or on macOS.
patch feature - Claude Code Releases sdkai ·
Claude-Code v2.1.202: Dynamic workflows, OpenTelemetry, and bug fixes
This update introduces a "Dynamic workflow size" setting to control agent count and adds OpenTelemetry attributes for better workflow tracing. Several bugs affecting session management, remote control, workflow parsing, and voice dictation have been fixed. These changes benefit users working with complex workflows, distributed tracing, and those encountering specific session or input issues.
patch - Anthropic News aiengineergovernment ·
Alberta Government Uses Claude to Enhance Cybersecurity
The Government of Alberta has successfully leveraged Claude Code with Opus and Sonnet models to review, identify, and fix cybersecurity vulnerabilities across millions of lines of government code. This initiative significantly accelerated the process of securing systems, which would have traditionally taken years. The program involved scanning 466 million lines of code in just 20 hours, remediating security gaps, and developing new tools to bolster system safety, showcasing a large-scale application of AI in government cybersecurity.
announcement feature - Claude Code Releases sdkaiengineer ·
Claude Sonnet 5 system role change for harness reminders
Claude Sonnet 5 sessions will no longer use the mid-conversation system role for harness reminders. This change impacts users who rely on this specific system role for reminder functionality. The update ensures cleaner and more direct communication within the API.
patch - Claude Code Releases sdkai ·
Claude Code v2.1.200: Dialogs, permissions, and background agent fixes
Claude Code v2.1.200 updates `AskUserQuestion` dialogs to require explicit opt-in for auto-continuation and changes the default permission mode to "Manual" across all interfaces for enhanced control. Several critical bugs affecting background session stability, agent startup, and data corruption have been resolved, ensuring more reliable agent operation. Additionally, improvements to UI rendering, accessibility, and the installation process are included for a smoother user experience.
patch - Anthropic News aisecurityengineer ·
Anthropic Details Fable 5 Cybersecurity Safeguards and Jailbreak Framework
Anthropic has redeployed Fable 5 globally, providing new details on its cybersecurity safeguards, specifically safety classifiers designed to detect and block dangerous uses. The company also introduced a draft AI jailbreak severity framework developed with Glasswing partners, aiming to standardize risk communication. These updates are relevant to all users, security researchers, and AI developers interested in responsible AI deployment and risk mitigation.
feature announcement - Claude Code Releases sdkaimedia ·
Claude Code v2.1.199: Improved error handling and agent stability
This release addresses several stability and error handling issues within Claude Code, particularly concerning background agents and subagent interactions. It improves how stacked skill invocations are handled and ensures partial responses are preserved during server errors. These fixes benefit developers using Claude Code, especially those running complex agent workflows or experiencing transient API issues.
patch - Anthropic Python SDK Releases sdkai ·
Anthropic SDK Python v0.115.1
Anthropic has released version 0.115.1 of its Python SDK. This patch primarily involves chore updates, specifically the removal of non-functional types from the SDKs. The update is available now and affects developers using the Python SDK.
patch - Claude Code Releases sdkaigaengineer ·
Claude in Chrome, Background Agents, and New Skills Go GA
Claude in Chrome is now generally available, bringing enhanced background agent notifications and new skills like `/dataviz` for chart design. Background agents have improved Git integration, committing and opening draft PRs upon completion. These updates are available to all users, with several fixes addressing session stability, UI rendering, and agent behavior across web, desktop, and VS Code environments.
feature patch - Claude Release Notes ai ·
Access to Claude Fable 5 and Claude Mythos 5 Suspended
Anthropic has suspended access to the Claude Fable 5 and Claude Mythos 5 models. The company aims to restore access as quickly as possible. Further details are available in a separate statement.
announcement - Claude Release Notes aipreviewengineer ·
Anthropic Enterprise: Model Entitlements Beta
Anthropic's Enterprise plans now offer a beta feature for admins to control user access to specific models and effort level settings. This enhancement allows for greater organizational oversight of AI model usage and configuration. The feature is currently in beta and is available to administrators on Enterprise plans.
feature announcement - Claude Platform Release Notes ai ·
Claude Platform Updates: New Models, Features, and Deprecations in July 2026
Anthropic has released Claude Sonnet 5, Claude Fable 5, and Claude Mythos 5, featuring expanded context windows and updated tokenizers, impacting prompt token counts. Several enhancements to Claude Managed Agents improve session event streaming, pagination, and configuration overrides. The platform also introduced new tool versions for code execution and web search, alongside API and SDK updates, while deprecating older Opus and Sonnet models and fast mode for certain versions.
feature patch deprecation - Anthropic News aigaengineer ·
Anthropic Fable 5 Model Access Restored After Export Control Lifted
Anthropic's Fable 5 and Mythos 5 models had their access restricted due to US export controls, which have now been lifted. Fable 5 will be globally available starting July 1st, with specific access terms for different plans and platforms. The company is also enhancing its safeguards and collaborating on an industry framework for AI model safety.
feature announcement - Claude Release Notes ai ·
Anthropic Launches Claude Sonnet 5
Anthropic has released Claude Sonnet 5, an improved version of their agentic Sonnet model. This update offers significant enhancements in reasoning, tool use, coding, and knowledge work compared to its predecessor, Sonnet 4.6. The new model is available now for users seeking more advanced AI capabilities.
feature announcement - Claude Platform Release Notes aigapreviewengineer ·
Claude Platform Updates: New Models, Rate Limits, and SDK Enhancements
Anthropic has released Claude Sonnet 5 and Claude Fable 5, introducing new pricing and a 1M token context window. Rate limits have been standardized across models, and several older models are retired. The release also includes SDK updates, new tool capabilities, and enhancements to managed agents, affecting developers and architects working with the Claude API.
feature deprecation patch - Claude Platform Release Notes aigadeprecationengineer ·
Claude Sonnet 5 Launched with 1M Token Context, New Tokenizer
Anthropic has launched Claude Sonnet 5, featuring a 1 million token context window and a new tokenizer that increases token count by approximately 30%. This release impacts users migrating from Sonnet 4.6 due to changes in adaptive thinking and the removal of manual extended thinking. Sonnet 5 is available at introductory pricing until August 31, 2026, with Priority Tier support not included.
feature deprecation patch
About Anthropic release tracking on ReleaseBytes
Anthropic ships Claude model releases, API updates and SDK versions on a rapid cadence. ReleaseBytes tracks the official Anthropic announcement channels and SDK changelogs, summarising each release and flagging model deprecations and breaking changes so teams building on Claude stay ahead of them.
Frequently asked questions
How often are Anthropic release notes updated on ReleaseBytes? ›
Continuously. ReleaseBytes monitors the official Anthropic release channels around the clock and publishes a plain-English summary of each announcement shortly after it lands.
What kinds of Anthropic 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 Anthropic releases? ›
Set up a free email or Slack alert filtered to Anthropic, 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 Anthropic release data come from? ›
From the official sources: Anthropic news, Claude releases and SDK updates. Every item links back to the original vendor announcement.