xAI Release Notes

Follow

305 release notes curated from 99 sources by the Releasebot Team. Last updated: Nov 11, 2026

Get this feed:

xAI Products

  • Nov 5, 2026
    • Date parsed from source:
      Nov 5, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.77

    Grok Build fixes terminal input, copying, MCP reconnections, context usage, and resume-list behavior.

    • [ui] kitty_keyboard = "query" lets Shift+Enter insert newlines in unidentified terminals
    • Quote copy now preserves & in links, and double-clicks no longer fold thoughts when missing text
    • MCP server connections no longer restart unnecessarily on session reloads or updates
    • Ctrl+L now works on more screens, and copy hints match what actually copies to the clipboard
    • Context usage shown in the footer no longer drops to zero or shows stale values after a rewind
    • Rows in the /resume list now expand and collapse the same way they do inside a session
    Original source
  • Nov 4, 2026
    • Date parsed from source:
      Nov 4, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.76

    Grok Build improves terminal screen redraws after clearing and correctly handles shell startup functions with special characters.

    Ctrl+L redraws the screen after a terminal clear without flashing

    Shell commands correctly handle startup functions containing special characters

    Original source
  • All of your release notes in one feed

    Join Releasebot and get updates from xAI and hundreds of other software products.

    Create account
  • Nov 4, 2026
    • Date parsed from source:
      Nov 4, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.75

    Grok Build improves terminal usability and reliability with clearer permission warnings, smoother prompt and session controls, better workflow status, and fixes for rendering, exports, and configuration behavior across platforms.

    • Permission rules that do not load, such as a rule that does not parse or a .claude/settings.json that is not valid JSON, show a warning when a session opens instead of being skipped silently
    • Cmd+Up / Cmd+Down and Ctrl+Home / Ctrl+End move the cursor to the start or end of the whole prompt, and with Shift they select to there
    • Send now shows as Enter Enter in the footer where Ctrl+Enter arrives as plain Enter, such as inside tmux with its default settings
    • Clicking a link or file that cannot open, such as over SSH with no display, shows why, for example "Can't open links over SSH"
    • Left on an empty prompt opens the dashboard from any session or the welcome screen, not only from a session you opened from the dashboard
    • Workflow status lines say how many agents are still running instead of how many exist
    • Ctrl+L in the scrollback adds the selected message or highlighted text to the prompt as a quote; open Extensions with /plugins instead
    • Notes above a queued prompt, such as a model switch result or a ready worktree, stay on screen when the prompt starts instead of scrolling out of view
    • Stop buttons in the dock stop a background run while its approval card is open, instead of doing nothing until you answer the card
    • Rejecting a change keeps edits you saved to the file yourself, even when Grok missed the save, instead of overwriting them
    • grok --resume opens the session in the folder it was recorded in, wherever you launch Grok from, instead of refusing
    • Windows: on older consoles such as plain cmd.exe and Windows PowerShell windows, the ▶, ⏸ and ◌ marks in todo, workflow and list rows show as characters the console can draw instead of empty boxes
    • Generated image and video paths with wide characters such as Chinese, Japanese or emoji fit their row and keep the file name, and Windows paths keep it too
    • The rewind confirmation no longer crashes Grok on a short terminal or when you shrink the window while it is open
    • Footer and status bars in narrow terminals drop or shorten the items that do not fit, instead of painting them over each other or cutting buttons off
    • Cut-off Fetch, Web Search and MCP tool output ends with the same … +N lines · to view all row as code cells, naming your viewer key, and one hidden line reads +1 line
    • 16-color terminals show the paste label and tool accents in the muted gray other secondary text uses, and the paste label brackets no longer vanish on the light theme
    • A question card's Enter:submit button, the dashboard session popup's [✗], and collapsed workflow rows highlight under the mouse pointer like other clickable controls
    • Emoji and keycap symbols are measured at the width the terminal draws them, so rows no longer run past their box or lose their …
    • /export to a named pipe that nothing is reading fails right away instead of freezing Grok until it is killed
    • Typing a digit or symbol while the scrollback has focus, with vim keys off, moves focus to the prompt and types it there, as letters already do; ? still opens the command palette
    • grok sessions delete no longer fails with "Directory not empty" when another process is still writing into the session's folder
    • Dashboard rows stay in Working while a session's background subagents or workflow runs are going, and stopping the row ends them
    • Code blocks inside quotes render every line under the quote's bars, quote copies turn reference links into inline links, and lines after multi-line math are numbered correctly
    • Disabling a project's hook in /hooks applies only to that repository, instead of also skipping hooks with the same name in other repositories
    • Turning a skill on or off in the Skills tab saves only [skills].disabled to config.toml, so an organization's managed skill folders and ignore list keep applying
    Original source
  • Nov 2, 2026
    • Date parsed from source:
      Nov 2, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    xAI

    Introducing Grok Bot 1.0

    xAI releases Grok Bot 1.0, bringing shared AI teammates to teams with proactive task support, improved browser and app workflows, and stronger writing and presentation tools. Bots can join Slack, Microsoft Teams, and video calls, while new controls help teams manage access and usage.

    Bots now work with your whole team, take on work without being asked, and get more done, faster.

    Today we’re launching Grok Bot 1.0, a major update to our AI teammates. Since introducing Grok Bot in beta, we’ve made three major improvements:

    • Multiplayer teammates. You can share Bots and bring them into Slack, Microsoft Teams, and video calls. Bots learn from everyone they work with and keep improving their understanding of the team’s knowledge and preferences.
    • Proactive teammates. Bots can recognize work that needs to be done and offer to take it on. They can follow up on unfinished tasks, anticipate what a team needs next, and bring people in when a decision requires their judgment.
    • More capable teammates. Bots are faster and more reliable at using the browser and completing complex work, including writing documents and building presentations. We’ve also made them more efficient, reducing the resources required to complete tasks.

    Grok Bot began as a way to hand off work to an AI teammate. With 1.0, those teammates understand the work your team is doing and the people doing it, so they can contribute like any other colleague.

    Bots work together alongside your team

    With Team Bots, everyone on a team can work with the same AI teammate. Since we launched Team Bots, we’ve made a number of improvements to make them easier to share and manage. As a Bot works with different people, it learns more about the team’s work and how each person contributes, becoming more useful to everyone over time. Sharing is more flexible too. Share a Team Bot with specific people or groups, and work with it in Slack or Microsoft Teams.

    In conversations, Bots can follow discussions involving multiple people, understand who’s involved, and contribute relevant information or offer to pick up work. During a product launch, for example, a Bot might notice that new customer research points to a different message than the team is considering. It can bring that into the discussion, then coordinate with another Bot to update campaign materials once the team agrees on a direction.

    Bots can also join video calls and participate through the chat, bringing information from earlier conversations into the discussion. They keep track of decisions and learn more about the team’s work, allowing them to become more useful over time.

    Bots take initiative

    Grok Bot can spot work before anyone asks. It can monitor ongoing projects, recognize when something needs attention, and suggest next steps based on what it knows about the team’s goals, processes, and priorities.

    For example, a Bot helping with a product launch might notice an upcoming deadline and prepare the materials the team needs. It can also follow up on unfinished work, coordinate with other Bots, and bring people in when a decision requires their judgment.

    Bots learn and improve as they work

    As people work with a Team Bot, it builds an understanding of how the team operates. It learns individual preferences, remembers decisions, and draws on the expertise of different team members. When someone teaches a Bot a better way to complete a task, that improvement can benefit everyone who uses it. Bots can also suggest changes to how they work based on what they learn from the team.

    We’ve also added safeguards for working in shared spaces. Bots respect each person’s permissions, ask before taking sensitive actions, and check whether information is appropriate to share in a public channel. This allows teams to give Bots more responsibility while maintaining control over what they can access and do.

    Bots are smarter, faster, and more capable

    Beyond the changes to how Bots work with teams, we’ve continued improving the overall Bot experience. Some of the biggest changes since beta include:

    • Better browser use. Bots navigate websites more reliably, handle sign-ins more smoothly, and complete tasks with fewer interruptions.
    • Stronger writing and presentations. Bots produce better documents and slide decks, and can now create interactive Pages that you can edit together or ask them to keep updated.
    • More ways to work across apps. Bots have improved connections to tools like Gmail, Google Calendar, and GitHub, allowing them to complete more work directly.
    • Smarter approvals. Bots better understand when you’ve already authorized an action and when they need to ask, reducing unnecessary interruptions while protecting sensitive information.
    • New Bot settings and analytics. Manage what a Team Bot knows and see how your team uses it.

    Get started with Grok Bot 1.0

    Grok Bot 1.0 is available starting today. You can create a Bot for yourself, share it with your team, or bring it into Slack, Microsoft Teams, and your video calls.

    Team Bots are available on Team and Enterprise plans. Meetings are available on Team, Enterprise, and Ultra plans. Enterprise admins can turn on Meetings in settings.

    We’re excited to see what you build with your new teammates.

    Original source
  • Nov 2, 2026
    • Date parsed from source:
      Nov 2, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    xAI

    Giving every team a data scientist with Grok Bot

    xAI’s Grok Bot adds Data Bot, a shared AI teammate that helps teams query company data, run analyses, build dashboards, and track metrics in Slack, bringing data science expertise to everyday questions and workflows.

    Giving every team a data scientist with Grok Bot

    By Sally Xu

    Data Bot is a Team Bot in Grok Bot that the data science team teaches, so every team can query the warehouse, run analyses, and get answers the way a data scientist would.

    At SpaceXSI, our data science team works across the company, helping teams answer difficult questions and finding new ways to incorporate data into our products. Over time, we've developed an intuition for our data, including how to interpret it and where its limitations lie.

    But our small team cannot answer every question that arises in a fast-growing company. People come to us throughout the day with questions that may be straightforward but still require our expertise to answer accurately. This can easily make us a bottleneck.

    A few months back, we started exploring how AI agents could change that relationship. We wanted to give everyone a data scientist, available around the clock, that could draw on our collective knowledge, while shifting more of our own work toward building and improving that AI teammate.

    We built Data Bot, a Team Bot in Grok Bot that we've taught how our data science team works. It can query our warehouse, conduct analyses, and return results in the form each person prefers. Unlike personal bots, which depend on what their owners know to teach them, Data Bot draws on a shared set of skills that our data team maintains and improves for everyone.

    Teams can also add Data Bot to their Slack channels, where it answers questions and contributes proactively to ongoing work. As we improve its skills, everyone benefits from the same growing body of expertise.

    Teaching Data Bot how the data team works

    Data Bot needs enough context to approach a question the way our team would. We give it that context through skills, which cover these main areas:

    • Metrics and analysis. We define our core metrics across the business, such as active users and revenue, and identify the source tables and query logic behind them. The skills also guide Data Bot through investigating changes in those metrics, from breaking down a spike or drop by segment or region to checking for holidays or logging changes before attributing it to a product change.
    • Experiments. These cover how we set up A/B tests, run power analyses, check on exposure balance and sample bias, plan rollouts, deep dive into unintuitive results, and write up readouts once experiments have reached power.
    • Charts and dashboards. We equip Data Bot with data visualization best practices, such as using line charts for trends over time and horizontal bars to compare categories, along with our team's chart themes and style guides. Other skills guide Data Bot through creating dashboards, so it can turn a chart in chat into a Databricks or Hex dashboard the whole team can use.
    • Data engineering. These cover our best practices for writing and reviewing dbt models, our conventions for automated pipeline alerts, and how to add analytics event logging so we can measure how people use new features from the moment they launch.

    Data Bot learns from experience

    We also give Data Bot shared memory so it can retain knowledge that emerges through day-to-day use. For example, it might learn which teammate owns a particular pipeline or that a previous drop in an important metric turned out to be a regional holiday. The next time it investigates a similar change, it can draw on that experience and check the holiday calendar before attributing the drop to a product change.

    As the Team Bot's managers, we can review these proposed additions to its memory and approve or reject them before Data Bot shares that knowledge with everyone who uses it.

    Helping PMs better understand their products

    Product managers have a constant stream of questions about how people use their products, many of which fall outside what their existing dashboards can tell them. Answering those questions has traditionally meant digging into the data themselves or asking someone on the data team for help.

    Data Bot gives PMs a way to explore those questions as they arise.

    Kevin Niparko, a PM at SpaceXSI, can message it in Grok Bot much as he would a teammate, get an answer in minutes, and follow up without starting a new analysis. He describes the experience as having a data scientist in his pocket.

    The ease of asking means PMs can investigate many more questions throughout the day. Previously, even opening a dashboard or starting an analysis in Cursor introduced enough friction to discourage some questions. Now Kevin can explore things he might otherwise have let go, without having to ask a data scientist for help.

    Data Bot also adapts to how people prefer to consume information. Kevin likes a concise headline accompanied by a chart, so Data Bot presents its findings in that format, letting him move quickly from a question to an answer he can use.

    Helping marketers understand what their campaigns accomplish

    Marketing teams need to understand what customers do after receiving a campaign. Email platforms track opens and clicks, but those metrics don't always show whether the campaign achieved its goal.

    Jon Brenner, who works on lifecycle marketing at SpaceXSI, uses Data Bot as an on-call analyst. He can give it a Braze campaign ID and ask what recipients did afterward. Data Bot queries the warehouse to see whether people who clicked went on to use the feature.

    For our recent launch of The X Subscription, Jon wanted to know how many email recipients actually switched plans. Braze's built-in conversion metric couldn't answer that question, so Data Bot queried the warehouse to track how many recipients received the email and how many went on to switch plans.

    Katy Abbaszadeh, our director of revenue growth, uses Data Bot to measure the impact of events on sales. She tracks how events contribute to pipeline generation and help accelerate existing deals. She also uses Data Bot to monitor weekly progress against targets, investigate what might be slowing progress, and turn the results into charts for presentations.

    With Data Bot, our marketing team can now answer more of their own questions without waiting for the data team, while still drawing on the team's expertise.

    Letting Data Bot handle entire analytics workflows

    Data Bot can take on work that goes well beyond answering individual questions. The data team has given it the tools to conduct more substantial analyses, including projects that would ordinarily require several rounds of collaboration with a data scientist.

    A PM investigating a change in product adoption, for example, can ask Data Bot to work out what happened. Data Bot can plan the analysis, query the warehouse, and explore possible explanations. It can then produce a written report with charts or build a dashboard in Hex that the team can continue using.

    Data Bot joins team channels and contributes proactively

    Teams can add Data Bot to their Slack channels and give it responsibility for ongoing reporting. For example, a team can ask it to track an important metric and post weekly updates, following the reporting conventions our data science team has established.

    Data Bot can run that workflow on a schedule and contribute updates directly to the group conversation. Colleagues can then ask follow-up questions or have it investigate an unexpected change, much as they would with another teammate.

    Data science expertise for all

    Data Bot adoption is growing quickly within SpaceXSI. Last week alone, we estimate it handled about 70 hours of work that would otherwise have landed on the data science team.

    We still handle complex analyses that require more judgment, and we're learning where Data Bot needs additional guidance. Our goal is to give colleagues across the company an AI teammate who can draw on our expertise whenever they need it, while leaving the Data Science team more time to focus on the questions that need our direct attention.

    Original source
  • Similar to xAI with recent updates:

  • Nov 2, 2026
    • Date parsed from source:
      Nov 2, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.74

    Grok Build improves its terminal experience with clearer workflows, more reliable session handling and safer updates across platforms. It also fixes prompt freezes, reduces CPU and memory use, and adds smoother file, link and subagent interactions.

    Messages between subagents appear as left-aligned rows under the bullet that name the sender and recipient, replacing the centered rows

    Shift+Enter inserts a newline in VS Code and other xterm.js-based terminals that use the Kitty keyboard protocol

    /mcps: when a browser cannot open an MCP server's sign-in URL, the URL shows in a panel you can copy with c or the [copy the url] button

    Shell approval cards say when Auto-review blocked a command or could not review it, and Ctrl+F shows the reviewer's explanation

    ! mode Tab completion offers commands installed as symlinks, such as Homebrew, npm, rustup and Debian alternatives binaries

    Quotes in replies get a [Copy] button that copies the whole quote as Markdown without its > markers

    Named subagent links open on a plain click, so they work over SSH from macOS where a modified click cannot reach Grok

    @ paths outside the workspace: an @ path that starts with /, ~/, or ../ lists files from that folder; set [ui] at_picker_outside_workspace = false to turn it off

    Picked @ files in the prompt can be edited in place: Backspace at the end of one reopens the picker on its path, and Left or Right puts the cursor inside it to change the path

    Built-in skills, agents and workflows are refreshed when their local cache's manifest is missing or corrupt, instead of staying outdated or failing every sync

    Ctrl+Y on a question card no longer closes it without an answer; while you type an answer, it pastes back text cut with Ctrl+K

    /new tabs show the context window you just switched to before the new session replies

    Resumed sessions show a background command's completion in the turn that ran it, with its real run time instead of about 0.0s

    grok update still runs when --sandbox or config sets a sandbox profile, unless Grok is already running confined

    A code block [Copy] in the transcript copies while a queued prompt is open for editing, instead of copying nothing or closing the edit

    grok logout stays signed out when another running Grok refreshes its sign-in, instead of signing you back in on the next launch

    Opening links and files from the TUI no longer leaves a zombie process behind for each click

    A failed session delete shows the "Could not delete session" prefix once, so the reason is no longer cut off

    Windows: Grok exits fully after a startup error, a /minimal or /fullscreen switch, or a signal, instead of sometimes hanging

    Grok running confined in a sandbox never starts an update on its own and tells you to run grok update instead

    Quitting to install an update with Ctrl+U on the welcome screen exits within the usual exit deadline instead of sometimes hanging after "Update installed"

    grok update --force-reinstall, or grok update --version with the installed version, keeps your working grok on Linux and macOS when the download fails its checks, instead of deleting it

    Windows installers check that the downloaded build runs before replacing grok.exe and agent.exe, and keep the existing install when it does not, such as when a proxy serves a blocked-download page

    Command-line arguments that are not valid UTF-8 no longer crash grok at startup: paths such as --cwd accept them, and text such as -p gets a usage error

    Installers keep the installed grok when a proxy or cache returns an empty download, instead of replacing it with an empty file and reporting success

    Environment variables with non-UTF-8 bytes no longer stop Grok from starting

    Workflow run lines in the transcript end with what the run returned instead of a bare colon, /tasks shows the time it was taken, and polling an unknown task id leaves a row with its error

    Headless runs stopped by Ctrl-C or SIGTERM end their json, streaming-json or streaming-messages-json output with a closing record that keeps the run's outcome, instead of cutting off the stream

    macOS: the TUI no longer freezes when it runs inside a sandbox that blocks the window server

    grok -p --resume shows why a session could not load instead of always "Session does not exist", and with --output-format json a resume, continue or fork that fails before the session opens prints one JSON error

    Sessions with an unrecognized reasoning effort or unreadable bytes in their history resume instead of failing: the session runs at the model's default effort and says so once, and the unreadable lines are left out of the transcript

    Windows: on older consoles such as plain cmd.exe and Windows PowerShell windows, status marks and selection arrows in workflow views, the dashboard, and the feedback and extensions dialogs show as >, •, √ or x instead of empty boxes

    File paths with spaces in inline code, such as ~/Library/Application Support/..., link to the whole path when it exists instead of stopping at the first space

    Feedback that cannot be sent leads with the outcome, such as "Could not send feedback. Saved to Drafts.", then gives the reason and says to retry from /feedback

    Dialog tab labels, shortcuts and settings rows that do not fit end in … instead of being cut off

    Prompt history search highlights matched letters in the same color as other fuzzy lists, and the counts on the / and @ menu borders are dimmed like other border captions

    Dashboard: the peek header reads Responding while a reply streams, and the Esc footer hint reads new agent like the other lowercase hints

    Custom agent files saved with a UTF-8 byte order mark, as some Windows editors write them, load instead of being skipped

    Running markers in workflow views and MCP servers that are still starting use the theme's running color, like other progress indicators

    Minimal mode: "Sign-in failed" shows in the error color instead of the warning color

    A larger context window picked with /context-window stays in effect, so compaction no longer starts early after the session sits idle, after a failed request such as a rate limit, or until you restart

    Long sessions use less memory: finished Markdown blocks far off screen release their rendered layout, and search still finds their text

    Typing no longer lags when ffmpeg is not installed and PATH holds slow directories, such as Windows drives under WSL

    Long pastes that arrive as keystrokes, such as pastes on native Windows or in terminals without bracketed paste, no longer freeze the prompt for seconds while the plan-mode tip is on

    Pasting long single-line text such as logs, JSON or base64 into the prompt no longer freezes the TUI for seconds

    File watching on Linux no longer keeps a CPU core busy for about a minute after startup in large repositories

    Lower CPU use while subagents run: the TUI repaints only for animations on screen

    grok --resume opens much faster on large session stores, reading only the resumed session instead of every session

    Original source
  • Nov 2, 2026
    • Date parsed from source:
      Nov 2, 2026
    • First seen by Releasebot:
      Nov 5, 2026
    xAI logo

    xAI

    grok-imagine-image-quality retired

    xAI routes Grok Imagine image requests to a lower-priced 2.0 low-quality model without requiring request changes.

    Requests to grok-imagine-image-quality and grok-imagine-image-pro are now served by grok-imagine-image-2.0-low: grok-imagine-image-2.0 with quality set to low, at a lower per-image price. No request changes are needed.

    usage.cost_in_usd_ticks on each response reflects the lower price, and gRPC responses (xAI SDK) report grok-imagine-image-2.0-low as the model. See the migration guide.

    Original source
  • Oct 28, 2026
    • Date parsed from source:
      Oct 28, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.73

    Grok Build releases a broad usability and reliability update with improved picker navigation, a Gruvbox theme, multiline sessions, and Windows fixes. It also strengthens update and code-restore handling while polishing transcript, image queue, and terminal behavior.

    • Page Up / Page Down, Home and End move the selection in /skills and other pickers
    • Automatic updates stop retrying a new version that fails to start on this machine until a newer version is out, and each launch says to run grok update to see why
    • Gruvbox dark theme (gruvbox-dark-hard), selectable with /theme
    • Start in multiline in /settings ([ui] multiline_default) starts new sessions and the dashboard in multiline mode, where Enter inserts a newline
    • Windows: quitting Grok ends grok.exe once the terminal is restored, so the shell no longer sometimes waits with no prompt
    • Several Send nows in a row no longer label the turns they stop "Turn cancelled by user" or jump the transcript view
    • --restore-code: when the code restore fails, your uncommitted and untracked files stay in the working tree instead of in a stash
    • Windows: shortcuts typed with Shift-held symbols or AZERTY digit keys work
    • Queued images show as green chips in the queue pane, as in the prompt
    • Typing while the transcript has focus moves any character, including non-ASCII text, to the prompt, and leaves a pending card or plan approval as it is
    • Worktree creation notes stay on screen when a queued prompt starts
    • Session load errors show the whole reason instead of cutting out the middle of a long message
    • Lines with | that are not a Markdown table wrap like other text instead of being cut off at the terminal edge
    • Transcript selection stays on the same row when a row above it is removed
    • Kitty graphics protocol: images Grok hides are freed in the terminal, so they no longer use its memory
    • --restore-code restores its own stash even when another worktree changes the stash list meanwhile, and keeps your changes stashed when a post-checkout hook fails after the checkout
    Original source
  • Oct 27, 2026
    • Date parsed from source:
      Oct 27, 2026
    • First seen by Releasebot:
      Nov 11, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.72

    Grok Build ships workflow and interface improvements, including clickable file and pull request links, safer session deletion, and fixes for custom models, MCP tool results, and long server replies.

    • Relative file paths in the agent's output, such as src/app/main.rs, open on click when the file exists under the session's folder, as absolute paths already do
    • Subagent and background task ids mentioned in the transcript open that subagent's view or the task's output on click
    • Pull-request numbers such as #123 in the agent's replies open that pull request on the session repository's forge
    • Subagents named in a reply show as the subagent's label instead of its raw id once the reply finishes, as a link that opens its transcript
    • /workflow runs detail view: → or a click focuses a phase's agent list, ↑ / ↓ pick an agent, Enter opens it, and ← returns to the phases
    • Feedback draft rows show on their own instead of inside a group of tool rows, and opening one shows the /feedback Drafts tab with that draft selected
    • Warp: clicking a URL that wraps across rows opens the whole URL instead of only its first row
    • /delete confirmation starts on Cancel, so pressing Enter no longer deletes the session; press 1 to delete
    • Max thoughts width (/settings, [ui].max_thoughts_width) sets how wide thinking blocks wrap; it previously had no effect
    • Dashboard shows the new folder on the dispatch box's top border right after a Ctrl+L or /cd pick
    • Custom models on the Responses API with a custom base_url no longer fail every turn with an HTTP 400 schema error for use_tool
    • /feedback Drafts tab keeps the selected draft when the list reloads, such as after leaving Drafts and coming back
    • MCP tools built with FastMCP or the official MCP SDK send their result to the model once instead of twice, so each call uses about half the context
    • Plan review notes accept every key, so a note that starts with a no longer approves the plan
    • MCP servers over stdio no longer lose part of a long reply line, such as a large tool list, when a read is interrupted
    Original source
  • Oct 27, 2026
    • Date parsed from source:
      Oct 27, 2026
    • First seen by Releasebot:
      Nov 2, 2026
    xAI logo

    xAI

    Custom moderation controls

    xAI now lets API users set content guidelines for Grok responses across NSFW, violence, and profanity categories.

    You can now set content guidelines on text API requests. Add a guidelines array to a Chat Completions or Responses API request to set how Grok handles content in the nsfw, violence, and profanity categories. Grok follows the guidelines as it writes its response. See Custom Moderation Controls.

    Original source
  • Oct 23, 2026
    • Date parsed from source:
      Oct 23, 2026
    • First seen by Releasebot:
      Oct 26, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.70

    Grok Build now separates answered questions with blank lines outside compact mode and reports loaded plugin skills and those skipped for external links.

    Answered questions now have a blank line between them, except in compact mode (/compact-mode)

    • /plugins now counts only the plugin skills that load, and says how many were skipped because they link outside the plugin folder
    Original source
  • Oct 23, 2026
    • Date parsed from source:
      Oct 23, 2026
    • First seen by Releasebot:
      Oct 26, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.69

    Grok Build improves dashboard navigation with folder-based agent grouping, clearer branch and folder labels, and session resume across folders. It also updates welcome-menu shortcuts, retires an unused setting, improves config diagnostics, and returns users to the shell faster when quitting.

    • Dashboard directory grouping now shows a header with the folder path and agent count above each folder of agents
    • Dashboard location (branch and folder) now sits on the dispatch box top border instead of its own header row
    • /resume on the dashboard now lists sessions from every folder, with the current folder and its checkouts first
    • Welcome menu now lists Resume session (Ctrl+R) above New worktree (Ctrl+W)
    • [ui] simple_mode is retired and no longer has any effect; sessions always open with the prompt focused
    • grok doctor and /doctor now point out a leftover [ui] vim_mode key (now [keys] vim) or the retired [ui] simple_mode key in your config
    • Quitting Grok returns you to your shell sooner
    Original source
  • Oct 22, 2026
    • Date parsed from source:
      Oct 22, 2026
    • First seen by Releasebot:
      Oct 26, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.68

    Grok Build improves worktree reliability and session continuity with clearer copy failures, preserved model context, and accurate image-send status. It also streamlines Vim navigation settings and refines terminal themes, mode notices, and dashboard controls.

    • Worktree create and fork now list files that could not be copied and any mode changes that were lost
    • Vim navigation is now one setting, [keys] vim (off or nav), also set with /vim-mode; the older [ui] vim_mode still works
    • Terminal theme now draws a subtle band behind your prompts using the colors your terminal reports
    • Mode-switch toast now shows the mode name (Agent, Plan, Ask) in the color of that mode
    • Change-directory picker (Ctrl+L on the dashboard) opens with the path field unfocused, like the /resume search
    • Image placeholders in the transcript now show whether the actual image was sent or only the text was typed
    • Worktree create now fails cleanly instead of falling back when it cannot copy tracked files or runs out of disk space
    • Running Grok sessions are no longer broken when an update deletes their binary
    • Dashboard model picks now keep their selected context window when spawning the next session
    • Send Now prompts and rewound turns now correctly restore and mark any images that were originally sent
    • Resumed sessions now show the correct sent or struck-through state for image placeholders
    Original source
  • Oct 21, 2026
    • Date parsed from source:
      Oct 21, 2026
    • First seen by Releasebot:
      Oct 23, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.67

    Grok Build improves worktree approval handling, vim-mode prompt suggestions, and headless command exits. It also restores themes when leaving the picker without confirming and speeds up quitting sessions with hooks.

    Worktrees also use saved "always allow" answers from the main checkout, so you are not asked again for a tool you already allowed there. Set [features] worktree_inherits_main_approvals = false to turn this off; the main checkout "never allow" answers apply either way

    Prompt suggestions follow vim mode: with [ui].vim_mode on, Tab moves focus to the scrollback instead of accepting the suggestion. → or Ctrl+→ accepts a suggestion, and Esc dismisses it

    Headless grok -p runs exit when their work is done instead of hanging when a background task or subagent finishes and its output is read in the same turn

    The theme picker in Settings restores your previous theme when you leave it any way other than Enter, as /theme already does

    Quitting a session with hooks configured is faster

    Original source
  • Oct 20, 2026
    • Date parsed from source:
      Oct 20, 2026
    • First seen by Releasebot:
      Oct 23, 2026
    xAI logo

    Grok Build by xAI

    Grok Build 1.0.66

    Grok Build restores themes when leaving the picker, keeps selected rows readable across themes, and speeds up file completion results.

    Theme Picker

    /theme restores your previous theme when you leave the picker any way other than Enter, such as Esc, Ctrl+U, Tab then Esc, or opening Settings or the dashboard

    Readability

    Selected and highlighted rows stay readable on every theme because reverse-video text is never dimmed. This fixes faint descriptions in selected menu rows on the terminal theme in terminals such as Kitty

    File Completion

    @ file completion shows results faster after you type

    Original source
Releasebot

Curated by the Releasebot team

Releasebot is an aggregator of official release notes from hundreds of software vendors and thousands of sources.

Our editorial process involves the manual review and audit of release notes procured with the help of automated systems.