Grok 4.5 Lands as an 'Opus-Class' Model, and xAI Is Now Part of SpaceX

Anything with Elon Musk's name on it tends to arrive wrapped in noise, and his AI efforts are no exception. Earlier this year I wrote about Grok being briefly banned in Malaysia over deepfake misuse, then restored once safeguards were in place. This time the news is about the product and the company behind it, and it is genuinely significant. So let me do the opposite of the usual hype: stick to what reputable outlets have reported, and pull out the one lesson that matters for a business owner here in Sabah.

What actually happened

On 8 July 2026, xAI released Grok 4.5, its new flagship model built for coding, agentic tasks and knowledge work. Musk described it plainly: 'It is an Opus-class model, but faster, more token-efficient and lower cost,' and said the company's internal assessment puts Grok 4.5 as roughly comparable to Anthropic's Claude Opus 4.7, just much faster. In simple terms, xAI is claiming it has closed a lot of the gap with the top labs, while being cheaper to run. The company says Grok 4.5 is about twice as token-efficient as other leading models, and it launched with pricing of two US dollars per million input tokens and six per million output.

What makes this release unusual is who built it. Grok 4.5 was developed jointly with Cursor, the popular AI coding tool, and it is available inside Cursor and xAI's own Grok Build environment. That partnership is not a coincidence, and it leads straight into the bigger story.

It is an Opus-class model, but faster, more token-efficient and lower cost.from Elon Musk, on Grok 4.5

The corporate shake-up behind the model

Here is the part that surprised even people who follow this closely. Earlier in 2026, xAI merged into SpaceX, folding Musk's AI company into his rocket company. Then, in June 2026, SpaceX formalised an all-stock deal to acquire Cursor for an implied value of around sixty billion US dollars, a transaction reported by CNBC, TechCrunch, Forbes and Bloomberg and expected to close later in the year. The logic, as reported, is that Cursor had been held back by a lack of computing power and will now train its models on xAI's Colossus supercomputer in Memphis, one of the largest AI training clusters in the world.

You do not need to track the corporate chess to take the point. The AI industry is consolidating into a handful of enormous, deeply funded players, and the lines between rocket companies, coding tools and chatbots are blurring. Grok 4.5 is the first visible product of that reshuffle. Whatever you make of Musk, the scale of money and compute now sitting behind these tools is not small.

Why this matters for Sabah businesses

You might be thinking: I run a clinic, a kedai, or a tour business in Sabah, why should a new Grok model and a sixty-billion-dollar merger concern me at all. Fair question. Here is the connection. More of us are starting to lean on AI for real work, and the choices are multiplying fast. A year ago the practical decision was mostly ChatGPT versus a couple of others. Now Grok is pitching itself as a genuine, cheaper rival to the tools you may already use. That is good for you, because competition drives quality up and prices down. But it also means more options to weigh, and more marketing noise to see through.

It is worth remembering the earlier Grok chapter too. This is the same tool Malaysia briefly blocked over deepfake abuse before restoring it once safeguards were added. A model getting faster and cheaper does not, by itself, tell you the company behind it handles safety and your data the way you would want. Capability and trustworthiness are two different questions, and you have to ask both.

The practical takeaway

So what do you actually do with this. My advice is the same one I give in every workshop: choose your AI tools the way you would choose a business partner. Before you build a habit around any tool, ask three plain questions. Is the company behind it taking safety and your data seriously. Does it have a track record of behaving responsibly. And does it fit the actual job you need done, rather than just being the loudest name in the headlines this week.

• Pick tools for the job, not for the hype. A cheaper, faster model is only better if it also fits your work and you trust the people running it.
• Do not scatter your data across every new tool. Every place you paste customer information is a place you have to trust. Fewer, more reputable tools is safer than chasing each launch.
• Judge track record, not just benchmarks. A model can top a leaderboard and still have a history of being banned somewhere for misuse. Both facts count.
• Stick to stable providers for anything important, so a sudden change at one company does not put your daily work at risk.

None of this means you should be afraid of AI, or of Grok specifically. The opposite. The businesses that learn to use these tools well, and choose them wisely, will pull ahead of the ones that either ignore AI or use it carelessly. The Grok 4.5 news is just a useful reminder that judgement matters as much as capability.

That is exactly the kind of thing I teach in my hands-on AI workshops here in Sabah. Not just which buttons to press, but how to compare tools, protect your data, and use AI in a way that is both productive and responsible for a real Malaysian business. If you would like a practical, jargon-free place to start, take a look at the workshop below.