Skip to content
Back

AI posts

AI release notes

Official updates from Anthropic, OpenAI, Google, xAI, Mistral, DeepSeek and Meta, newest first. Open any update for a short AI summary and the full notes.

Compiled by the Toolstash team · Updated

  1. Mistral AI · Mistral API

    OCR 4.0 is deprecated and retires on September 30, 2026

    Use OCR 4.1 (mistral-ocr-4-1 or mistral-ocr-latest) instead, at the same price. Leanstral 1.5 (labs-leanstral-1-5), an experimental Labs model, is deprecated and retires on September 30, 2026. Z.ai GLM 5.2 (zai-glm-5-2) is deprecated and retires on October 31, 2026.

  2. Anthropic · Claude Code

    Claude Code 2.1.284

    14 new57 fixes29 other

    Added Claude Sonnet 5.5 (claude-sonnet-5-5), now the default Sonnet model on the Anthropic API — 1M context, $2/$10 per Mtok with $0.20/Mtok cache reads. Added a "Yes, but ask again next time" answer to auto mode's prompt before a read outside the working directories, so you can allow that one read and still be asked about later ones.

  3. Anthropic · Claude API

    Code written for Claude Sonnet 5 can break on Claude Sonnet 5.5 in five ways

    To turn off up-front thinking, send thinking: {"type": "between_tools"} instead of "disabled", at high effort or below. Forced tool use (tool_choice types any and tool) returns a 400 error. Thinking blocks are tied to the model and the conversation. On the Claude API and Google Cloud, the earlier computer_20251124 computer use tool isn’t accepted. The advisor tool rejects Claude Opus 4.8, Claude Opus 4.7, and Claude Sonnet 5 as advisors. For model-specific prompting patterns, see Prompting Claude Sonnet 5.5. Thinking blocks that Claude Sonnet 5.5 produces work only in the account that produced them, or in an account linked to it.

  4. Anthropic · Claude appsAnnouncement

    Claude Sonnet 5.5 launch

    We just launched Claude Sonnet 5.5, the second model in our Claude 5.5 family. Sonnet 5.5 is a faster, lower-cost complement to Claude Opus 5.5.

  5. Anthropic · New modelAnnouncement

    New model: Claude Sonnet 5.5

    We’ve launched Claude Sonnet 5.5 (claude-sonnet-5-5). It’s available on the Claude API, Claude in Amazon Bedrock, Claude Platform on AWS, Claude on Google Cloud, and Claude in Microsoft Foundry. For its context window, output limits, and prices, see the Claude Sonnet 5.5 model page.

  6. OpenAI · Codex CLI

    Codex CLI 0.158.0

    5 new6 fixes

    Configure copy-on-select and right-click paste in the fullscreen TUI. Copied transcript selections now preserve Markdown formatting. Connect to MCP servers that require pre-registered OAuth client secrets, including through codex mcp add --oauth-client-secret.

  7. Mistral AI · New model

    Z.ai GLM 5.3 is now Generally Available

    Z.ai GLM 5.3 (zai-glm-5-3) is now Generally Available.

  8. OpenAI · Codex CLI

    Codex CLI 0.157.1

    1 other

    Release highlights could not be determined: the supplied PR index is empty, and the GitHub tag comparison returned 404.

  9. Anthropic · Claude Code

    Claude Code 2.1.283

    11 new53 fixes30 other

    Added x-claude-code-prompt-id to the gateway hint headers so LLM gateways can group the requests that serve one user prompt; opt in with CLAUDE_CODE_GATEWAY_HINT_HEADERS=1. Added availableModelsMatch managed setting: with "exact", an availableModels entry allows only the model version it names, so new releases stay blocked until listed.

  10. Anthropic · Claude appsAnnouncement

    Build plugins for Claude

    You can now submit plugins to the Claude directory through a new developer portal, track them through review, and see usage analytics once they’re live.

  11. OpenAI · Codex app

    macOS security update

    Fixed CVE-2026-100754 on macOS in version 26.924.20706, with thanks to Patrick Wardle, Objective-See Foundation.

  12. OpenAI · Codex CLI

    Codex CLI 0.157.0

    6 new6 fixes

    Added GPT-6 Sol and Luna, including Amazon Bedrock support and migration prompts for older models. Enabled fullscreen transcripts by default and added Shift-click to extend text selections.

  13. Anthropic · Claude Code

    Claude Code 2.1.282

    8 new58 fixes20 other

    Added a maxProseWidth setting that caps the width of Claude's prose in wide terminals while tables and code blocks keep the full width. Added a startup notice, and /status and claude doctor entries, listing telemetry variables in a project's settings files that were ignored or that turned telemetry off.

  14. Anthropic · Claude API

    We’re resuming billing for refusals that arrive before any output when stop_details.category is "bio",…

    Mid-stream refusals were already billed. Refusals billed under this change are charged like any other request, at the rates of the model that ran it. Refusals before any output in other categories are still not billed, and fallback credit is unchanged. This change applies on all platforms. The Compliance API local session endpoints are out of beta for Claude for Microsoft 365 sessions in Excel, PowerPoint, Word, and Outlook (product_surface values beginning with office_agents). The Compliance API Activity Feed no longer returns file names, project document names, or artifact titles.

  15. Google · Gemini CLI

    Gemini CLI 0.61.0

    5 fixes3 other

    preserve explicit versioned Flash model IDs. prevent indirect prompt injection via build file modifications and untrusted flags.

  16. Anthropic · Claude Code

    Claude Code 2.1.281

    13 new112 fixes51 other

    Added Claude apps gateway support for newer Claude Desktop keys in desktop policy blocks, including blockReadsOutsideWorkingDirectories and disableBypassPermissionsMode. Added assume_role on Claude apps gateway Bedrock upstreams: the gateway calls Bedrock as an IAM role it assumes through STS, in another AWS account if needed, optionally one session per developer.

  17. Anthropic · Claude API

    Cache diagnostics is out of beta on the Claude API and no longer requires the cache-diagnosis-2026-04-07…

    Include the diagnostics object on a Messages request to opt in; requests that still send the header work as before. Responses from POST /v1/messages now always include the diagnostics field, which is null when the request did not include the diagnostics object.

  18. OpenAI · Codex app

    ChatGPT for iOS

    Redesigned home screen. New iPad split view keeps your task list beside the open task in landscape. Added support for browsing changes inside nested Git repositories. Side chats now carry over attachments, selected text, and review comments. Fixed replies appearing blank or incomplete when the next queued prompt starts. New tasks preserve your selected computer and better recognize project checkouts and worktrees. More reliable file links, with clearer loading feedback when opening files and new tasks. Clearer error messages explain why task history could not load and what to try next. Smoother animations when adding photos from the camera or Photo Library. Fixed pairing failures caused by an incorrect device clock and SSH connection issues on Mac and Linux.

  19. OpenAI · Codex CLI

    Codex CLI 0.156.1

    1 new

    Choose GPT-6 Sol or GPT-6 Luna from the model picker. The rate-limit switch prompt now recommends GPT-6 Luna.

  20. Google · Antigravity

    Direct subagent messaging syntax, Vim numeric count multipliers, headless lifecycle handling, and…

    Introduces direct @ prompt syntax with interactive autocomplete, adds Vim count multipliers across Normal and Visual modes, enhances headless execution lifecycle and task timeout handling, and fixes file-lock concurrency issues in conversation history databases along with streaming response stability fixes.

  21. Anthropic · Claude Code

    Claude Code 2.1.280

    12 new67 fixes35 other

    Added Claude Opus 5.5 (claude-opus-5-5), now the default Opus model — 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads. Added mouse support to more lists in fullscreen mode: the wheel scrolls the /skills list, and a skill's state options in /plugin can be clicked.

  22. Anthropic · Claude API

    On Claude Opus 5.5, thinking can’t be disabled: thinking: {"type": "disabled"} and thinking: {"type":…

    Omit the thinking field and control thinking depth with the effort parameter. tool_choice types any and tool also return a 400 error, as on Claude Fable 5.1; use auto with strict tool use. On the Claude API and Google Cloud, computer use on this model requires the computer_toolset_20260801 toolset and the earlier computer_20251124 tool returns a 400 error; on Amazon Bedrock, computer_20251124 keeps working. Fast mode (research preview) is available for Claude Opus 5.5 on the Claude API. Tools can now be defined inside a mid-conversation system message, in beta on the Claude API with the inline-tools-2026-09-15 beta header.

  23. Anthropic · Claude appsAnnouncement

    Claude Opus 5.5 launch

    We just launched Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 on most work and costs 40% less to run than Opus 5.

  24. Anthropic · New modelAnnouncement

    New model: Claude Opus 5.5

    We’ve launched Claude Opus 5.5 (claude-opus-5-5), a model for long-running agentic coding and knowledge work. It has a 1M token context window by default, 128k max output tokens, and always-on adaptive thinking, at $4 / $20 USD per MTok (Claude Opus 5 is $5 / $25 USD). Claude Opus 5.5 is available on the Claude API, Claude in Amazon Bedrock, Claude Platform on AWS, Claude on Google Cloud, and Claude in Microsoft Foundry.

  25. OpenAI · Codex app

    Codex app

    Added a notification sound setting for task completions, permission requests, and questions. Choose Default, Classic, or None. On macOS, you can also choose a system sound or import an audio file. Performance improvements and bug fixes. Added a sign-in recovery dialog for session errors that prevent app startup. Preserved chat titles in the Rename dialog when worktree setup finishes.

  26. OpenAI · Codex CLI

    Codex CLI 0.156.0

    6 new6 fixes3 other

    Choose an optional fullscreen UI with /tui for your next launch, including transcript search, mouse selection, and right-click copying. Voice conversations are enabled by default, with an F8 toggle, a /voice settings picker, and bundled audio runtimes for Linux and Windows.

  27. OpenAI · New model

    GPT-6 Sol and Luna in Codex and ChatGPT Work

    GPT-6 Sol and GPT-6 Luna are rolling out to Codex and ChatGPT Work at lower token prices than their GPT-5.6 predecessors. Use Sol for complex coding and agentic workflows, and Luna for focused, high-volume tasks. Both models are rolling out to Plus, Pro, Business, Enterprise, and Edu users. Free and Go users can access Luna in the desktop app. Availability depends on rollout and workspace settings; Enterprise administrators must enable the new models. In ChatGPT, the models are available in Work and Codex, but not Chat. Choose a model in the picker when available. In the CLI, use /model or start a task with codex --model gpt-6-sol or codex --model gpt-6-luna.

  28. Google · Antigravity

    Proportional context compaction budgets, full-window sizing, custom model media support, and leak fixes

    Improves context compaction by spreading user-request budgets proportionally across all prompts and sizing truncation ceilings from the model's full context window, enables default PDF and audio support for custom models, routes Ctrl+G editor launches to a real terminal stdout, caps voice dictation recordings to avoid deadline failures, and resolves resource and memory leaks across conversations and background pollers.

  29. Google · Antigravity

    Windows Subsystem for Linux, Office attachments, and live subagent cards

    Connect to a Windows Subsystem for Linux distribution from Application Settings, and drop Word, Excel, and PowerPoint files straight into a prompt. Subagents now appear as live cards you can follow and stop while they work. This release also adds right-click copy and save for images.

  30. Google · Antigravity

    Plan before you build

    Type /plan and the agent drafts a plan you can read and edit before it writes any code. The Agent setting that governs this is now called Plan Review Policy, and you can set it to review every plan, review only when the agent judges it worthwhile, or skip review entirely.