Grok Build

The decision comes down to two things: where you already live and how much effort you are willing to put into access.

Who it is for. For those already using Grok and wanting to try vibecoding without new subscriptions: the build button is included with every plan, and the finished app immediately lives at a link. Also suitable for engineers looking for a cheap terminal agent: in our measurement, a task cost $2.59 versus $11.75 at Claude Code with nearly the same score (measurement 16.07.2026).

Who it is not for. For those not ready for preparation steps from Russia: you will need a foreign IP address and card. Also not for those who value tool maturity: the terminal version has been in early beta since May 2026, and practitioners note rough error handling.

Starting from scratch? The chat door is exactly for this: you do not need to write code at all. But to understand what to ask the agent for and how to check the result, start with the page about vibecoding — the link is below.

A good fit - section 3 shows both doors. Choosing between it and Claude Code: the decision is in section 7.

What vibecoding isStarting from scratch: how to set tasks for agents with no coding experience

One agent, two faces: in chat it builds whole apps, in the terminal it works like a regular code agent.

In chat. Describe a game, website, or dashboard in words, and Grok builds a working version right in the conversation: edits are in words too, iteration after iteration. The finished app is published to a grok.me address with one button, and you choose visibility, from «only me» to «the whole internet». In the official demo, the Hyper Racer arcade game is built in 12 minutes 4 seconds.

You can remake someone else's. Any published app can be forked with the remix button and becomes your draft. Dashboards connect live data through connectors, while an integrated secrets manager stores third-party service keys.

The code is not locked in. The whole project is exported to a GitHub repository: then you can continue in your usual editor or the terminal version of the agent. You can attach your own domain to the published app.

In the terminal. A full-screen text interface, written in Rust and fully open. Its signature move is parallel subagents, each in an isolated copy of the repository, so they do not trample each other's files. The rest is standard for the genre: headless mode for scripts, MCP, and custom project rules in AGENTS.md.

There are two doors: chat for everyone and the terminal for those who live in code.

You can use it from Russia. The first step is standard. Change your Russian IP address to a foreign one — American, Dutch, or any other. The vendor cites export restrictions, so without this the site may not open. Registration is easy after that: Google, Apple ID, or email, with no phone number required. Subscription payment is covered in section 5.

Chat. Open grok.com, switch to Build mode, and write what to build. It works in the browser, on iOS, and on Android. The apps are hidden from RU stores, so the browser is more reliable on a phone.

How to install the terminal version

Step 1. Install the agent with one command:

curl -fsSL https://x.ai/cli/install.sh | bash

Step 2. Sign in: SuperGrok and X Premium Plus subscriptions provide access, while an API key for API work is obtained at console.x.ai. For readers in Russia, the working route is SuperGrok, because X Premium Plus runs into X itself being unavailable.

Step 3. Give your first task in words and watch the process: the agent shows its steps and edits right in the interface.

Installation command verified 23.08.2026 against x.ai/news/grok-build-cli

An agent measurement is always a measurement of the pair «agent + model». An agent has no score on its own.

On the board. The Grok Build + Grok 4.5 pair holds 76,4 AA points (snapshot 16.07.2026) — right next to the best Claude Code pair with its 77,2. At the time of measurement, the agent was one and a half months past its announcement.

Price from the same measurement. In the run, the task cost $2.59 and took 16,5 minutes versus $11.75 and 23,5 minutes for the leader. Grok Build wins not on score but on price: almost the same result for four and a half times less.

Community measurement. 25,9 thousand GitHub stars in the first forty days of the public repository (measurement 23.08.2026). Stars do not measure code quality, but they clearly show how many engineers came to look.

Grok Build has no separate price: it is included in Grok plans, so you pay for the plan, not the agent.

Free plan. Since 19.08.2026, building in chat is available on every plan, including the free one. The vendor does not publish exactly how many requests it provides — there is no official number.

Paid tiers. SuperGrok costs about $30 per month, SuperGrok Heavy about $300. The terminal version is included with SuperGrok and X Premium Plus, but the X route is a dead end for readers in Russia, so consider the SuperGrok subscription the CLI's starting price.

Paying from Russia. Russian cards do not work: payment processing rejects them at the gate. The usual routes are a foreign card or an intermediary with a fee. For how to set it up, see payment guide.

We look at three things: where your project lives, what is visible in the agent's code, and how mature it is.

The project lives with the vendor. The chat version builds and publishes apps in the SpaceXAI cloud: the grok.me address lives on its servers. There is an exit: the whole project is exported to GitHub, so the work cannot be locked inside the service. An integrated secrets manager stores third-party API keys, and they are not written into code.

You can read the agent's code, but not fix it. The repository is open under Apache 2.0, and the agent is written in Rust. This is a storefront export: the vendor does not publish development history or accept external changes. It is enough to check «what the agent does on my machine», but not for community life.

Honestly young. The terminal version has been in early beta since 25.05.2026, and practitioners note that its error handling is rough compared with more mature agents. An unfamiliar agent gets its first tasks in a test project, not a production repository - that is user discipline with any agent.

The most common search question about Grok Build. Short answer: it trades maturity for price.

A different trade-off. Claude Code has the best pair of measurements (77,2 vs. 76,4) and a clear maturity advantage: a robust tool with an ecosystem built around it. Grok Build answers with price - in the 16.07.2026 measurement, the same task cost $2.59 versus $11.75, four and a half times cheaper with a gap of less than a point.

Different entry barriers. Claude Code lives in the terminal and expects you to read code. Grok Build has a door with no threshold: chat-based building on any plan, including the free one. For someone with no coding experience, it is the only one of the two paths open right away.

Who else is nearby. Model freedom is offered by OpenCode and aider: both are open and work with any provider, while Grok Build runs on its own grok-4.6.

The product is three months old. Its history is short but dense: from a closed beta for subscribers to open source and a button on the free plan.