SpaceX launched Grok 4.5, a new AI model optimized for coding and agentic tasks, priced at $2 per million input tokens and $6 per million output tokens—underpricing Anthropic's Claude Opus.
The model is immediately available through SpaceXAI's developer tools and follows SpaceX's acquisition of xAI in February, which is now becoming SpaceXAI.
The company is targeting the enterprise AI tools market, buoyed by its pending all-stock acquisition of Cursor-maker Anysphere for $60 billion（約9.6兆円）.
What happened
SpaceX on Wednesday launched Grok 4.5, an AI model trained across tens of thousands of Nvidia GB300 graphics processing units, immediately available through SpaceXAI's coding agent Grok Build, Cursor, and the SpaceXAI console via API key.
Why it matters
Grok 4.5 is priced at $2 per million input tokens and $6 per million output tokens—significantly cheaper than Anthropic's Claude Opus at $5 and $25 per million respectively—positioning SpaceXAI to compete for enterprise AI tools customers as Elon Musk integrates the xAI startup he acquired in February into SpaceX operations.
What to watch
EU availability is expected in mid-July; Musk claims Grok 4.5 is faster and more token-efficient than competing Opus-class models, though OpenAI is launching its own advanced model GPT-5.6 Luna on Thursday at $1 per million input tokens and $6 per million output tokens.
Ask the AI about this article →
SpaceX's launch of Grok 4.5 represents a consolidation of Elon Musk's AI ambitions under the SpaceX umbrella. The company acquired xAI in February and announced in May that xAI would cease to exist as a separate company and become SpaceXAI instead. That transition is now materializing through product launches, with Grok 4.5 arriving as a pricing alternative to established players in the enterprise coding-agent market.
The timing is aggressive: SpaceX is currently pursuing an all-stock acquisition of Anysphere, the startup behind Cursor (a popular AI coding agent), for $60 billion（約9.6兆円）. By launching Grok 4.5 through Cursor itself—partnering with the target acquisition—SpaceX is signaling that the combined entity will have its own frontier model at competitive pricing. Grok 4.5's $2/$6 input-output token structure undercuts Anthropic's Claude Opus ($5/$25) and matches OpenAI's GPT-5.6 Luna on the output side, though Luna is cheaper on input. The model was trained across tens of thousands of Nvidia GB300 graphics processing units with emphasis on data filtering and quality scoring, matching claims other frontier labs make about training discipline. EU availability within mid-July suggests SpaceX is moving quickly to expand geographic reach before competing launches mature.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Visko raised $10 million in pre-seed funding from Llama Ventures and opened public access to its first foundat…
U.S. markets ended August higher, with the S&P 500 up 2.6% and the Nasdaq up 3.9%
Neurovia AI, an Abu Dhabi-based company, is pitching Saudi security agencies software that it says can compres…
AI company Runway has unveiled Solaris, the first model in a new category it calls "Interface World Models." I…
Google's AI search gave advice to call emergency services for users alone with an African, Indian, or Pakistan…
John Deere introduced JD, a conversational AI tool that lets farmers ask open-ended questions about their hist…
