Grok Bot on Desktop and iOS: What's Actually Shipped

For a product built around the idea of an agent that keeps working after you close the app, where that app actually runs matters more than usual. Grok Bot's August 11 beta shipped on desktop and iOS, per xAI's own announcement, with enterprise access gated behind a waitlist rather than general availability. That's the confirmed baseline. The platform picture beneath it has a bit more texture, and a few gaps worth naming plainly.

Desktop: the full spread, with one detail worth flagging

layer3labs specifies desktop support across macOS, Windows, and Linux — a genuinely broad spread for a beta product, and one that includes Linux, a platform many consumer AI tools skip entirely. That breadth makes sense given Grok Bot's distribution through Cursor, a developer tool with a Linux-using audience that a consumer-only chat app wouldn't necessarily prioritize.

Desktop is also, per the product's own design, the more natural home for the browser-driven, filesystem-adjacent automation the earlier piece in this series examined — see Grok Bot's Cloud Computer for the specifics and the caveats around exactly what access level is confirmed versus reported. A desktop app is the expected front door for a tool whose core loop involves recording workflows, reviewing multi-step task output, and managing a growing library of taught routines — the kind of dense, file-and-window-heavy interaction that's awkward on a phone screen.

Mobile: iOS confirmed, Android still pending

The mobile story is less settled. xAI's own materials confirm iOS at launch. The Android question gets slightly different answers depending on the source: ayautomate describes Android as "coming soon," while layer3labs says it's simply "planned," a somewhat softer commitment. Neither phrasing comes with a date. Seven weeks after the initial beta launch — and by the time of this writing, roughly seven weeks past the business-tier expansion too — Android support for Grok Bot had not shipped in any source checked for this piece.

That's a real gap for a product positioning itself as an always-on teammate you might want to check in on from your phone throughout the day. It's also not unusual: mobile parity frequently lags desktop launches for agent products still finding their footing, and prioritizing iOS first is a common (if sometimes controversial) choice given the platform's typically higher-spending user base. Still, any organization standing up parallel bots per department with a workforce that skews Android should treat mobile access as a real constraint on rollout today, not a rounding error.

A tool built to work while you're away from your desk is only as useful as the device you actually have in your pocket. For Android users, that promise is currently "coming soon," not "here."

What "actually shipped" leaves out

It's worth being precise about a distinction this series returns to often: platform availability and feature completeness are different claims. Being available on desktop and iOS confirms you can install and open Grok Bot on those devices. It says nothing about whether every advertised capability — the "teach-a-task" recording mode covered in From Routine to Habit, for instance — is fully rolled out to every user on every platform simultaneously. layer3labs explicitly describes that recording feature as rolling out gradually, which suggests platform launch and feature completeness are being staged separately rather than shipped as one bundle. That's a reasonable engineering choice for a beta product, but it means "available on iOS" and "fully functional on iOS" aren't necessarily the same statement, and the launch materials don't give a clean way to tell which capabilities have actually reached which platforms yet.

How this compares to the competition's rollout maps

OpenAI's Dots launched reachable through ChatGPT, Codex, Slack, and Microsoft Teams, with phone/text access and an audio model both explicitly described as "coming soon" rather than shipped — a similar pattern of a broader platform ambition than what's actually available on day one. Meta's Muse rolled out on iOS, Android, and muse.ai in the US, with AI-glasses integration also pending — notably, Muse actually beat Grok Bot to Android, despite launching almost a month later. That's worth sitting with: xAI moved first into the always-on-agent category by nearly two months, but on raw platform coverage, Meta's month-younger product currently covers more ground.

None of the three products has finished rolling out to every platform and every promised feature as of this week — which is, in one sense, the whole point of this series existing right now. All three are genuinely mid-launch, and "what's actually shipped" is a moving target that's worth re-checking rather than assuming stays fixed from the announcement date. The next piece in this series, Beta Status, Real Stakes, looks at what xAI's own "evolving capabilities" language signals about how much of this is likely to change in the weeks ahead.

What platform gaps mean for a real rollout decision

For anyone actually deciding whether to deploy Grok Bot inside a specific team right now, platform coverage is a more concrete filter than most of the abstract safety questions this series raises elsewhere — it's the one variable you can check against your own organization's device mix today, with no ambiguity about the answer. A team that's entirely Mac and iPhone hits no platform friction at all. A team with a meaningful share of Android phones or Chromebook-only setups is, as of this writing, locked out of the mobile half of the product entirely, regardless of how the credential model or approval gates eventually shake out. That's worth checking first, before spending time on the harder questions about what a bot is allowed to do once it's running — there's no point evaluating Grok Bot's safety posture for a workforce it can't actually reach yet on the devices they carry.

It's also worth watching the sequencing of what ships next. If Android support lands before a more detailed safety framework does, that tells you something about xAI's priorities in this beta phase — reach before rigor. If the order runs the other way, that's a different signal entirely, and one worth revisiting as this series' later pieces track how the product matures.

Part 7 of 30 in the Grok Bot series. Previously: SuperGrok, Cursor, and the Quota Question. Next: Beta Status, Real Stakes: What Evolving Capabilities Means. Browse more builder insights or explore AI skills at aiskill.market.