Grok Bot just validated everything we've been building, at 10x our price. An honest comparison from a tiny competitor.

Two days ago xAI launched Grok Bot, always-on AI agents, each with its own cloud computer, that keep working after you close your laptop. It's distributed with Cursor and bundled into SuperGrok Heavy ($300/mo), Cursor Ultra ($200/mo), and Cursor Teams Premium ($120/seat). I'm the co-founder of Vestra, a 7-person team building an autonomous AI workspace in the exact same category. So yes, I have a horse in this race, full bias disclosed upfront. But I spent yesterday going through every launch review and walkthrough I could find and I think the honest comparison is more interesting than a marketing take. Including the parts where they're ahead of us. Where Grok Bot is genuinely impressive (credit where due): Teach-a-task is brilliant. You screen-record yourself doing a workflow once, and the bot learns it and repeats it independently. One reviewer taught a bot to pull newsletter stats with zero API, zero plugin, just by watching. This kills the biggest cost in automation: specifying the workflow. We don't have this yet. Full mobile parity. Every bot, routine, and live agent screen carries over to your phone, including taking over a session mid-task. Most agent products (ours included) are web-first. This is a real distribution advantage. The login handoff pattern. The bot navigates to a login page itself, hands control to you for the sensitive part, then resumes. Where I think they got it wrong (and where we deliberately went the other way): No model choice. At all. Grok Bot picks the model for every task with no override, no advanced mode — xAI won't even say which models the router uses. We built Vestra model-agnostic from day one: any frontier model, auto-selected or pinned. When a new model drops, it's an upgrade, not a migration. If you're locked to one lab's models, every task inherits that lab's bad days. Cloud lock-in. Grok Bot runs on xAI/Cursor infrastructure, full stop. Vestra runs on our cloud, your cloud, or on-prem. For anyone with compliance requirements, this isn't a nice-to-have. The price gate. Cheapest access to Grok Bot is $120/seat/month, individuals start at $200–300, no free tier. Vestra starts at $20/month. I get it, compute is expensive. But if the pitch is "delegate your busywork," pricing it above most people's busywork budget is a strange launch choice. It's built for developers, sold through a code editor. The distribution deal with Cursor means the audience is people who already pay for AI dev tooling. We're building for the other buyer: the operator, the founder, the agency owner who doesn't want to know what an MCP server is. They just want the invoice chased and the report done. Full honesty about where we stand: The thing I actually feel after this launch isn't fear, it's validation. When xAI, OpenAI, and Anthropic all converge on "agents with their own computers that own outcomes," the category is real. The open question is whether it gets won by whoever bundles it into a $200 subscription or by whoever makes it accessible to the millions of small teams doing the busywork by hand. Genuinely curious what this community thinks: if you've tried Grok Bot this week, what was your experience? And what would an agent need to do before you'd trust it with real work?