Features

• Long-running agents

  SpaceXAI

  Stays with multi-step work: research, analysis, codebase edits, and turning a product idea into a working first version.

  Grok 4.6agentsmulti-agentcomputer-use

• Self-testing and verification

  SpaceXAI

  On longer trajectories, checks its own work before moving on.

  Grok 4.6agentscoding

• Interactive and visual work

  SpaceXAI

  Stronger first passes on visual and interactive projects — structure and visual language in one pass.

  Grok 4.6uicoding

• 500k context window

  SpaceXAI

  Text and image inputs with text-only output and no text output limit.

  Grok 4.6apilong-context

• Reasoning effort

  SpaceXAI

  Configurable reasoning: low, medium, high (default), and xhigh.

  Grok 4.6apireasoning

• API pricing

  SpaceXAI

  $2 / $0.50 / $6 per 1M tokens (input / cached / output) below 200k prompt tokens; $4 / $1 / $12 above. Fast variant is twice the price.

  Grok 4.6apipricing

• Where it runs

  SpaceXAI

  Cursor, Grok Build, xAI API, OpenRouter, Vercel, Cloudflare, Amazon Bedrock, GitHub Copilot, Gemini Enterprise Agent Platform.

  Grok 4.6apiavailability

• Speed and efficiency

  SpaceXAI

  Twice as fast as comparable models at half the price, per the Grok 4.7 announcement.

  Grok 4.7modelcoding

• 500k context window

  SpaceXAI

  Text and image inputs with text-only output and no text output limit.

  Grok 4.7apilong-context

• Reasoning effort

  SpaceXAI

  Configurable reasoning: low, medium, high (default), and xhigh.

  Grok 4.7apireasoning

• API pricing

  SpaceXAI

  $2 / $0.50 / $6 per 1M tokens (input / cached / output) below 200k prompt tokens; $4 / $1 / $12 above. Grok 4.7 Fast is billed at twice those rates in Cursor and Grok Build.

  Grok 4.7apipricing

• Responses API reasoning payload

  SpaceXAI

  On the Responses API, grok-4.7 always returns reasoning.encrypted_content even when include omits it.

  Grok 4.7api

• Where it runs

  SpaceXAI

  xAI API (grok-4.7), US regional endpoint, Cursor, and Grok Build (Grok 4.7 Fast).

  Grok 4.7apiavailability

• Own cloud computer

  SpaceXAI

  Each Bot runs on a persistent cloud VM with a browser, filesystem, and terminal. All Bots on an account share that user-scoped computer (files, browser sessions, logins), with a separate screen per Bot.

  Grok Botcomputer-useagents

• Always-on runtime

  SpaceXAI

  Keep working 24/7 in the cloud after the client is closed.

  Grok Botalways-onagents

• Colleague messaging

  SpaceXAI

  Message a bot like a teammate. Bots remember conversations, share context in threads, and can coordinate in group chat.

  Grok Botmulti-agentmessaging

• Skills

  SpaceXAI

  A skill is a reusable set of instructions for how to do a task. Skills are available across Bots. Teach a task can record a browser workflow (up to ten minutes) into a draft skill when that control is visible.

  Grok Botskillsautomation

• Routines

  SpaceXAI

  A routine tells one Bot when to run a workflow — on a schedule or, where supported, after an event. Background routines can run while the laptop is closed. A Bot can own up to 50 routines.

  Grok Botautomationagents

• Supported platforms

  SpaceXAI

  Desktop app on macOS (Apple silicon and Intel) and Windows (x64 and Arm64); mobile app on iPhone (iOS 18+) and Android (Android 9+). Linux desktop and iPad are not supported.

  Grok Botavailability

• Approvals

  SpaceXAI

  Bots only pull you in for judgment calls. They follow up on dropped threads and nudge stalled handoffs.

  Grok Botagents

• Connectors

  SpaceXAI

  Sign into the tools you already use and work across apps and inboxes, including platforms with no clean API or MCP.

  Grok Botconnectorscomputer-use

• Plan mode

  SpaceXAI

  The agent explores and drafts a plan before edits. Only the session plan file may be edited until you approve; bash is not gated. Auto and always-approve do not skip plan review. Enter with /plan or Shift+Tab; the agent can enter plan mode on its own. Reopen with /view-plan.

  Grok Buildcodingagents

• Auto and always-approve

  SpaceXAI

  Ask (default) prompts for anything not already allowed. Auto uses a classifier to auto-approve safe tools. Always-approve skips permission prompts; deny rules and PreToolUse hooks still apply. Shift+Tab cycles Normal → Plan → Auto → Always-approve. Allow/deny rules in [permission] or --allow/--deny; deny wins.

  Grok Buildcliagents

• Headless and scripting

  SpaceXAI

  Run without the TUI via grok -p. Optional --output-format streaming-json for scripts, automations, or other apps.

  Grok Buildcliautomation

• Agent Client Protocol

  SpaceXAI

  Use Grok Build through the Agent Client Protocol (ACP) in other apps.

  Grok Buildcliagents

• Subagents

  SpaceXAI

  Independent child sessions with their own context; they return a summary to the parent. Built-in types: general-purpose, explore (read/list/search only), and plan (no shell, no edits). Add types under .grok/agents/ or ~/.grok/agents/. Personas are behavioral overlays via /personas.

  Grok Buildmulti-agentcoding

• Skills

  SpaceXAI

  Reusable folders with markdown, scripts, and resources (SKILL.md). Discovered from ./.grok/skills/, ~/.grok/skills/, plugin skills/, and [skills] paths. User-invocable skills appear as /<skill-name>. Capture a session as a skill with /skillify.

  Grok Buildskillscoding

• Plugins

  SpaceXAI

  Plugins add skills, agents, hooks, MCP servers, and LSP servers from ./.grok/plugins/, ~/.grok/plugins/, marketplace installs, [plugins] paths, or --plugin-dir. The TUI extensions modal opens from /plugins, /hooks, /skills, or /mcps.

  Grok Buildskillsmcpcli

• Marketplaces

  SpaceXAI

  TUI Marketplace tab installs plugins from [[marketplace.sources]] in ~/.grok/config.toml and ~/.grok/plugins/known_marketplaces.json.

  Grok Buildskillscli

• Claude Code compatibility

  SpaceXAI

  Reads Claude Code marketplaces, plugins, skills, MCPs, agents, hooks, and instruction files (CLAUDE.md, Claude.md, CLAUDE.local.md, .claude/rules/) alongside .grok/ with no extra setup.

  Grok Buildcodingskills

• Hooks

  SpaceXAI

  JSON hooks (command or HTTP) on lifecycle events. PreToolUse is the only blocking event (explicit deny, else fail-open). Personal ~/.grok/hooks/*.json; project .grok/hooks/*.json needs /hooks-trust or --trust. Also reads Claude Code and Cursor hook files.

  Grok Buildhookscoding

• MCP servers

  SpaceXAI

  MCP servers expose external tools as <server>__<tool>. Add with grok mcp add (stdio or HTTP/OAuth). Also declared in ~/.grok/config.toml or project --scope. Loads ~/.claude.json, .cursor/mcp.json, and .mcp.json unless compat is disabled. /mcps in the TUI.

  Grok Buildmcpconnectors

• AGENTS.md

  SpaceXAI

  Loads AGENTS.md, Agents.md, AGENT.md, CLAUDE.md, and *.md in .grok/rules/ (plus .claude/rules/ and .cursor/rules/) from repo root down to cwd; deeper files win. --rules appends session-only text. grok inspect lists loaded files.

  Grok Buildcoding

• Web and mobile

  SpaceXAI

  Grok Build is available beyond the terminal — web and mobile surfaces announced Aug 19, 2026.

  Grok Buildavailability

• Release channels

  SpaceXAI

  Installer respects GROK_CHANNEL=stable|alpha|enterprise (default stable). The welcome screen shows the active channel. Failed channel updates suggest the matching reinstall command. Persist the choice in ~/.grok/config.toml under [cli] channel.

  Grok Buildclialpha

• Sessions

  SpaceXAI

  Every conversation is saved under ~/.grok/sessions/. Resume with /resume, grok --resume, or grok -c. /fork branches a peer session; /rewind restores files and truncates the conversation. grok sessions list/search/delete and grok export manage them.

  Grok Buildcliagents

• Worktrees

  SpaceXAI

  Isolated git checkouts under ~/.grok/worktrees/<repo>/<name> so parallel agents do not overwrite each other. Start with grok -w, /fork --worktree, or Ctrl+W on the dashboard. Persist until grok worktree rm or gc.

  Grok Buildcligit

• Sandbox

  SpaceXAI

  Limits what an approved tool call can read, write, and reach on the network (Landlock on Linux, Seatbelt on macOS). Off by default. Profiles: off, workspace, devbox, read-only, strict. Enable with grok --sandbox, [sandbox] profile, or GROK_SANDBOX.

  Grok Buildcliinfra

• Agent Dashboard

  SpaceXAI

  Fullscreen overview of every session (needs input, working, idle, completed, failed). Open with Ctrl+\, /dashboard, or grok dashboard. Reply or answer permission prompts inline; Ctrl+W starts new agents in a worktree.

  Grok Buildcliagents

• Background tasks

  SpaceXAI

  Run commands, subagents, and monitors while the conversation continues (/tasks, Ctrl+G). /loop repeats a prompt on an interval (max 50, expire after 7 days). Monitors notify on each script line. Ctrl+; or /queue lists prompts queued behind the running turn.

  Grok Buildcliautomation

• Theming

  SpaceXAI

  TUI themes via /theme or [ui] theme in ~/.grok/config.toml. Built-ins include GrokNight (default), GrokDay, TokyoNight, RosePineMoon, and OscuraMidnight. theme = auto follows OS light/dark.

  Grok Buildcli

• Imagine Image 2.0

  SpaceXAI

  Quality Mode for precise generation and editing. API id grok-imagine-image-2.0. 1K and 2K stills.

  Grok Imagineimageapi

• First-class editing

  SpaceXAI

  Magic wand, segmentation, background removal, multi-ref editing (up to 5 inputs), smart resize across common aspect ratios.

  Grok Imagineimage

• Imagine Video 1.5

  SpaceXAI

  Text-to-video, image-to-video, and reference-to-video. Native 1080p for T2V and I2V (Jul 31, 2026).

  Grok Imaginevideoapi

• Imagine API

  SpaceXAI

  Generate and edit images and video through the xAI API. Image input billed separately from output.

  Grok Imagineapimedia

• Speech to speech

  SpaceXAI

  Realtime voice agents. Starting at $0.05 / min.

  Grok Voicevoiceapirealtime

• Text to speech

  SpaceXAI

  TTS at $15.00 / 1M characters.

  Grok Voicevoiceapi

• Speech to text

  SpaceXAI

  STT with adjustable VAD threshold (vad_threshold) for streaming and batch.

  Grok Voicevoiceapi

• Voice Think Fast 2.0

  SpaceXAI

  Introduced Jul 29, 2026. grok-voice-2.0 is deprecated; upgrade to grok-voice-3.0 or grok-voice-latest at the same price.

  Grok Voicevoice

• Grok Voice 3.0

  SpaceXAI

  grok-voice-3.0 for Speech to Speech; grok-voice-latest points to grok-voice-3.0. Same price as Grok Voice 2.0; API unchanged.

  Grok Voicevoiceapirealtime

• Grok Voice Transcribe 2.0

  SpaceXAI

  grok-voice-transcribe-2.0 on Speech to Text; grok-voice-transcribe-1.0 reached EOL Oct 2, 2026 and requests route to 2.0 at the same price.

  Grok Voicevoiceapi

• Custom voices

  SpaceXAI

  Clone a voice from a reference clip (max 120s) via /v1/custom-voices and use the returned voice_id for TTS and speech-to-speech.

  Grok Voicevoiceapi

• Flagship models

  SpaceXAI

  Grok 4.7 (grok-4.7) for agents, coding, and knowledge work. Text and image in, text out.

  Grok API / Consoleapimodel

• Media and voice

  SpaceXAI

  Imagine image/video generation and editing plus Voice realtime, TTS, and STT on the same API.

  Grok API / Consoleapimediavoice

• Console

  SpaceXAI

  Keys, playground, usage, and changelog at console.x.ai.

  Grok API / Consoleconsoleapi

• Partner surfaces

  SpaceXAI

  OpenRouter, Vercel, Cloudflare, Amazon Bedrock, GitHub Copilot.

  Grok API / Consoleavailabilityapi

• Subscriptions

  Cursor

  Monitor PRs, watch a Slack thread, or run scheduled tasks. Agents wake on events and can auto-subscribe to PRs they create.

  Cursor Cloud Agentsautomationalways-on

• /goal and /loop

  Cursor

  Assign a long-lived objective (e.g. fix flaky tests and make CI green). Pair with /loop for recurring check-ins.

  Cursor Cloud Agentsagentsautomation

• Subagents on their own machines

  Cursor

  Subagents run on isolated VMs with clean project copies. Swarm them for independent tests.

  Cursor Cloud Agentsmulti-agentcomputer-use

• Custom modes

  Cursor

  Pin skills as always-on modes so agents stay focused. Select from the / menu.

  Cursor Cloud Agentsskillsagents

• Steering

  Cursor

  Send a follow-up without interrupting the current turn. Press return twice or Send now.

  Cursor Cloud Agentsagents

• Builds

  Cursor

  Ready-to-use copies of the dev environment. Agents boot ~3× faster to first token; included with Cloud Agents.

  Cursor Cloud Agentsagentsinfra

• Conversation hooks

  Cursor

  Observe and control prompts, responses, thinking, subagents, compaction, and turn completion.

  Cursor Cloud Agentshooksagents

• Start from scratch

  Cursor

  Cloud Agents can start without a connected GitHub or other SCM. Cursor creates an Origin repo in the background; save with Create repo when ready.

  Cursor Cloud Agentsagentsgit

• Live preview

  Cursor

  Port-forwards the cloud agent environment to the browser for preview and design mode.

  Cursor Cloud Agentsagentscomputer-use

• Vercel publish

  Cursor

  Connect a Vercel account and publish the agent build to a live URL.

  Cursor Cloud Agentsagentsconnectors

• Self-hosted machines

  Cursor

  Run cloud agents on machines in your network. My Machines connects a laptop or VM; team pools scale workers on demand with hibernate and restore. Integrates with AWS Lambda, Coder, Cloudflare, Daytona, Modal, Namespace, Vercel, and E2B. Self-hosted workers support computer use on Linux and Mac.

  Cursor Cloud Agentsagentsinfracomputer-use

• Projects

  Cursor

  Long-running work with a coordinator agent that plans and delegates to implementing agents. Shared context files sync across cloud and local machines; runs on a cloud computer when the laptop is closed.

  Cursor Cloud Agentsagentsmulti-agentautomation

• Rollouts

  Cursor

  Change monitor on every pull request: writes a monitoring plan, tracks deploys per environment, and reports verified healthy, regression detected, or inconclusive. Connects Origin or GitHub, CD deploy events, and telemetry (Datadog and others).

  Cursor Cloud Agentsautomationagentsinfra

• Security Review

  Cursor

  Posts one review comment per pull request reporting exploitable bugs (injection, auth bypass, secrets in source, SSRF, unsafe deserialization, vulnerable dependencies). Team rules enforce codebase-specific policies. Draft PRs are skipped.

  Cursor Cloud Agentsautomationagents

• Agent SDK

  Cursor

  Agents as code: instructions.md, tools/, skills/, MCP connections, and channels. `agent-sdk dev` runs locally; `agent-sdk deploy` hosts on Cursor. Evals and `agent-sdk github replay` test PR-driven bots.

  Cursor Cloud Agentsagentsapiautomationskillsmcp

• Live apps

  Cursor

  Cloud agents scaffold web apps, provision Postgres and Cursor SSO auth, run security review, and publish to a chosen domain for team or public sharing.

  Cursor Cloud Agentsagentshostingautomation

• Origin repos

  Cursor

  Create and host repos from the Codebase tab. Clone and push with the CLI. URL form cursor.com/codebase/{name}.

  Cursor Origingithosting

• GitHub sync

  Cursor

  Bidirectional PR sync and real-time updates. Disconnect anytime.

  Cursor Origingit

• Pull requests

  Cursor

  Timeline, commits, checks, diffs, comments, and merges in Origin, two-way synced with GitHub.

  Cursor Origingitagents

• Agents in every repo

  Cursor

  Query code, make changes, update PRs, and push branches from hosted repos.

  Cursor Originagentsgit

• App extensions

  Cursor

  Origin App Marketplace lists 20+ partner apps (deployments, CI, code review, security, monitoring, PM). Public submissions with review. Connect from the Apps tab.

  Cursor Originconnectors

• Grok 4.7 Fast in Cursor

  Cursor

  Grok 4.7 Fast is available in Cursor at twice the public API token rates (per xAI API release notes).

  Cursor IDE / CLImodelapi

• Grok 4.6 in Cursor

  Cursor

  Flagship model available in Cursor on Aug 12, 2026 with a one-week 2× usage boost.

  Cursor IDE / CLImodelcoding

• CLI

  Cursor

  Clone and push Origin repos, drive agents, and script workflows from the terminal.

  Cursor IDE / CLIcligit

• Google Workspace plugins

  Cursor

  Gmail, Drive, and Calendar access from Cursor (Aug 3, 2026 changelog).

  Cursor IDE / CLIconnectors

• iPad app

  Cursor

  Inbox, PR reviews, and multi-PR sessions on iPad.

  Cursor IDE / CLIavailability

• Early Access and Nightly

  Cursor

  Pre-release desktop builds via Settings → Beta → Update Access. Feature changelog entries ship with the minor; patch builds are not listed separately.

  Cursor IDE / CLIalphaide

• X Activity API

  X

  Realtime user-activity webhooks. Successor path from legacy Account Activity API. Subscribe by user id; optional direction filter.

  X APIapiactivity

• Pay-per-usage credits

  X

  No subscription SKUs. Buy credits, set spending limits, earn up to 20% back in xAI API credits. Owned reads vs public lookup priced separately.

  X APIapipricing

• Webhook event billing

  X

  Activity API events billed per delivery, same 24h UTC dedup window. Inbound events cost; many outbound actions do not.

  X APIapipricingactivity

• Media upload

  X

  Chunked upload via v2 paths (initialize, append, finalize). Post-create size and duration limits follow the authenticated user's Premium/verified status and media_category.

  X APImediaapi

• Mute and block events

  X

  mute.mute / mute.unmute / block.block / block.unblock. Private events; OAuth 2.0 user-context with mute.read or block.read.

  X APIactivityapi

• Lead Generation API

  X

  X Ads API for Lead Gen campaigns: create and manage native lead forms, attach LEAD_FORM destination cards, and export submissions as JSON, CSV, or XLSX. Requires Campaign user permission; lead exports restricted to account administrators.

  X APIapi

89 of 89