xAI Rebrands as SpaceXAI, Preps Grok 4.5 Built With Cursor Tech

Elon Musk has closed another chapter in the integration of his companies. xAI, the maker of the Grok chatbot, has formally adopted the name SpaceXAI, completing the absorption into SpaceX that began in February. Days later, the companies are set to unveil their first jointly built model, developed with Cursor, the coding tool SpaceX is acquiring for $60 billion.

The new SpaceXAI logo blends the letters AI into SpaceX's signature symbol. This is more than cosmetic. The formal disappearance of the xAI brand signals that Musk now treats artificial intelligence as core to the rocket company's strategy, not a separate side project.

Why Rockets and AI

Musk justifies the merger by pointing to constraints on ground-based infrastructure. Data centers on Earth face lawsuits, resident protests, and power grid shortages. SpaceX sees space as a way to sidestep some of those barriers in the long run, though for now the priority is combining computing resources, the Starlink satellite network, and software into a single ecosystem.

Space is the only way to scale at scale - Elon Musk, SpaceX/SpaceXAI

In 2025, the company spent $12.7 billion on AI infrastructure, more than three times what it spent on space launches and Starlink combined. Those figures show where Musk's empire actually centers its weight today, even though the word Space still leads the name.

What Grok 4.5 Brings

Grok 4.5 is the first model SpaceXAI and Cursor built together, even before the acquisition formally closes. It isn't a fine-tuned version of the existing Grok, but an architecture built from scratch on the Colossus supercomputer, with Cursor's coding data folded in during supplementary training.

The model underwent private testing at SpaceX and Tesla, where engineers evaluated it in their day-to-day coding work. The company says feedback was positive enough that it decided to move up the public launch to July 9, one day after the rebrand was announced.

Early evals show performance close to, perhaps exceeding Opus - Elon Musk, SpaceX/SpaceXAI

Musk is comparing the results to Anthropic's Claude Opus 4.8, and industry analysts are also pitting Grok 4.5 against OpenAI's GPT-5.5. However, the cited tests were run by in-house engineers at SpaceX and Tesla, companies controlled by the same owner as the model's maker. No independent benchmark, such as LMArena, Artificial Analysis, or SWE-bench, has published its own evaluation of Grok 4.5 yet, and neither the official price nor the context window size has been disclosed.

The Battle for Developers

Michael Truell, Cursor's founder, teased a model meant to challenge 'the frontier models from Anthropic and OpenAI.' That's a clear signal SpaceXAI wants into the coding assistant market, long dominated by Claude Code, OpenAI's Codex, and rivals like Zhipu AI's GLM-5.2. Pairing Colossus's computing power with Cursor's user base and data is meant to give SpaceXAI an edge that xAI alone never had.

For Polish software companies and developers using AI-assisted coding tools, this means another serious contender to test alongside Claude, GPT, and Gemini. The coding assistant market keeps getting more crowded, and the real differentiator is shifting toward who has access to the best training data from real code repositories, not just the most computing power.

What Comes Next

The July 9 launch coincides with the broad rollout of OpenAI's GPT-5.6 series and pricing changes for Anthropic's Claude Fable 5, making July 2026 shape up as a month of intense price and technology competition among the top AI labs. A final verdict on Grok 4.5 will have to wait until independent benchmarks verify Musk's claims and Cursor users put the model through its paces on their own projects.

Sources: Gizmodo (gizmodo.com), The Information (theinformation.com), BigGo Finance (finance.biggo.com), ExplainX (explainx.ai)