Grok 4 is an AI model from xAI that powers the Grok assistant, with built-in tools such as web browsing and a code interpreter, plus the ability to search X. The detail people miss: xAI's docs now list newer Grok 4.x models, and live facts need search switched on.
Grok 4 is the AI model behind Grok, the assistant you'll have seen on X, and people search for it because the name keeps showing up next to bold claims. I find it helps to separate what the maker says about it from what it means for an ordinary business owner.
So here's a plain explanation of Grok 4: what it is, what xAI says it can do, how it's different from the newer Grok versions now listed in xAI's docs, and how I'd treat it if I were using it for work.
Grok 4 in plain words
Grok is the AI assistant from xAI, Elon Musk's AI company. Grok 4 is one model in that family, the engine that reads your question and writes the answer. xAI's own announcement calls it the most intelligent model in the world, which is a claim to weigh up rather than a fact to repeat.
According to Wikipedia, Grok 4 was released on July 9, 2025. Since then, xAI has kept releasing newer versions, so Grok 4 is better thought of as the start of a line than the latest word.
What xAI says Grok 4 can do
xAI's announcement describes several abilities that set Grok 4 apart from a plain chatbot:
- Native tool use: it can use tools such as a code interpreter and web browsing, and it picks its own search queries.
- Searching X: it can search X using keyword and semantic search, and it can view media.
- Grok 4 Heavy: a version that uses parallel test-time compute, which xAI describes as weighing multiple hypotheses at once.
- Voice with a camera: in voice mode you can point your camera at a scene and get real-time analysis.
For developers, xAI says the Grok 4 API offers text and vision understanding and a 256,000-token context window. A context window is how much text the model can keep in view at once, and I've explained the idea in plain terms in my other AI guides.
Grok 4 and the newer Grok versions
Here's where people get confused. xAI's models page now lists newer text models, including grok-4.7, grok-4.6, grok-4.5 and grok-4.3. It describes grok-4.7 as the most capable model xAI has built.
| Grok model name in xAI's docs | What the docs say |
|---|---|
| grok-4.7, grok-4.6, grok-4.5 | Newer models, each with a 500k context window |
| grok-4.3 | An earlier model in the same line |
| Grok 4 (announcement) | The model this page is about, with a 256,000-token context window in its API |
The docs list Grok 4.x names rather than a model simply called Grok 4, so if you're choosing a model today, read xAI's current models page rather than relying on any article, including this one.
Where Grok 4 gets its facts
xAI's docs make an important point about freshness. Real-time information needs server-side Web Search or X Search switched on. Without them, Grok has no knowledge beyond its training data.
That matters if you use Grok for market research or news. An answer about recent changes to a competitor's website might simply be out of date unless search was involved. Ask it to show where a claim came from and check the source.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Grok 4 compared with ChatGPT and Claude
Grok's big difference is its link to X: it can search posts there directly, which is handy for spotting what people are saying about a topic. For writing, planning and working through documents, the leading assistants are closer than their marketing suggests. My comparison of Grok and ChatGPT walks through that from a small business point of view, and my simple guide to using Grok covers the basics of the app.
Points to think about before relying on Grok 4
Wikipedia notes that Grok has drawn public criticism and regulatory attention over some of its outputs at different points. I'd read reputable coverage for yourself rather than relying on my summary, and I'd apply the same checks I use with any assistant:
- Review anything it writes before it goes to a customer or onto your website.
- Verify names, dates and claims against a primary source.
- Don't paste client secrets, passwords or API keys into any chatbot.
- Keep a human in charge of anything that touches your brand's public voice.
- Read xAI's privacy and data controls pages before you use it with business information.
Using Grok 4 in a small business
- Ask it to summarise what people on X are saying about a trend in your industry, then open the posts to check.
- Have it draft a few versions of a short post, then rewrite the best one in your own words.
- Use it to explain a technical document in plain English before a client call.
- Ask it for questions a customer might raise about your service, then answer them on your site.
- Compare its answer with a second assistant on anything important.
If you'd like a steer on which assistants suit your team and how to set them up safely, that's what I do as an AI consultant.
Frequently asked questions
What is Grok 4?
Grok 4 is an AI model from xAI that powers the Grok assistant. xAI describes it as using native tool use, including a code interpreter and web browsing, and searching X.
Who makes Grok 4?
xAI makes Grok. Wikipedia says Grok 4 was released on July 9, 2025.
Is there a newer version than Grok 4?
xAI's models page lists newer models including grok-4.7, grok-4.6, grok-4.5 and grok-4.3. Check xAI's docs for the current list.
What is Grok 4 Heavy?
xAI describes Grok 4 Heavy as a version that uses parallel test-time compute to weigh multiple hypotheses at once.
Does Grok know today's news?
xAI's docs say realtime data needs server-side Web Search or X Search. Without them, Grok has no knowledge beyond its training data.
