Key Takeaways:
- SpaceXAI released Grok 4.5 to the public on Tuesday, July 8
- Musk claims it matches Claude Opus 4.8 at lower cost and higher speed
- OpenAI counters with GPT-5.6 public launch set for Thursday
Key Takeaways:

SpaceXAI released Grok 4.5 to the public Tuesday, claiming Opus-class performance at lower cost in the first model built jointly with Cursor.
SpaceXAI released Grok 4.5 to the public Tuesday, claiming the model matches Anthropic's Claude Opus 4.8 on performance while running faster and at lower cost, in the first joint release since its $60 billion acquisition of Cursor.
"Based on strong positive feedback from customers in our beta test program, SpaceXAI will make Grok 4.5 available to the public tomorrow," Elon Musk, chief executive officer of SpaceXAI, said in a post on X. "It is an Opus-class model, but faster, more token-efficient and lower cost."
Grok 4.5 runs on xAI's 1.5 trillion-parameter V9 foundation, with supplemental training data from Cursor's coding platform. The model entered private beta at SpaceX and Tesla on June 28. Internally, SpaceXAI compared its performance to Claude Opus 4.8 and OpenAI's GPT-5.5, according to a report from the Information. The launch comes two days before OpenAI plans to release its GPT-5.6 family — Sol, Terra and Luna — to the public on Thursday.
The release intensifies the rivalry between Musk and OpenAI, the company he co-founded in 2015 and left in 2018. Grok 4.5 targets complex, long-running tasks across software engineering, legal services and financial analysis, with enhanced cybersecurity capabilities — a broader work portfolio than Cursor's previous coding-focused models. The timing pits the two companies' latest models against each other as developers and enterprises begin evaluating competing systems.
Grok 4.5's positioning as an Opus-class model places it just below Anthropic's most powerful offering, Claude Fable, which remains limited in availability. Musk said in June that the unreleased model's performance was comparable to "some unspecified version of Claude Opus," and the Information later reported the specific benchmark target was Opus 4.8. SpaceXAI did not disclose independent benchmark results or the test conditions for its performance claims.
The Cursor acquisition, announced June 16 for $60 billion, gives SpaceXAI access to Cursor's AI coding platform and its user base of software developers. Cursor founder Michael Truell announced last month that the startup was working on a model that would compete with frontier models from Anthropic and OpenAI, calling it the "next phase of the company." Grok 4.5 represents the first output of that combined effort.
OpenAI's countermove arrives Thursday with the public launch of GPT-5.6 Sol, Terra and Luna, expanding access from a small group of vetted partners. The company previewed the model family in late June. The back-to-back releases give enterprise customers a direct comparison point between the two ecosystems.
This article is for informational purposes only and does not constitute investment advice.