SpaceXAI has released Grok 4.5, its latest artificial intelligence model and the company’s first major model launch since it became publicly traded several weeks ago. The company said the new system is designed to handle a wide range of tasks, including coding, app development, office work, research, writing and other forms of routine knowledge work.
In a blog post published Wednesday, SpaceXAI described Grok 4.5 as a general-purpose workhorse intended to automate many of the tasks that have become central targets for the artificial intelligence industry. The company also highlighted the model’s claimed efficiency, saying it delivers “twice greater token efficiency” than other leading models.
SpaceXAI said the improved efficiency could reduce the cost of using advanced AI systems, as expenses associated with token usage have become a growing concern for users of large language models. The company released benchmark results that it said demonstrated Grok 4.5’s competitiveness with other leading models, while noting the model was just short of being the highest-performing system in those comparisons.
Founder Elon Musk promoted the model on his social media platform X, which is a subsidiary of SpaceXAI, describing Grok 4.5 as an “Opus-class model.” Musk compared the system to Opus, Anthropic’s large language model designed for complex and intensive tasks.
“Based on strong positive feedback from customers in our beta test program, @SpaceXAI will make Grok 4.5 available to the public tomorrow,” Musk wrote in his post on X. He said the model was faster, more token-efficient and lower cost compared with comparable systems.
Musk later said SpaceXAI’s internal assessment was that Grok 4.5 was roughly comparable to Opus 4.7 but operated at a higher speed. He said the combination of capability, speed and cost was the factor that made the model competitive.
SpaceXAI said Grok 4.5 would be priced at $2 per million input tokens and $6 per million output tokens. The company positioned the pricing as competitive if the model’s performance matched its stated capabilities.
By comparison, Opus 4.7 was listed as costing $5 per million input tokens and $25 per million output tokens. Other artificial intelligence providers have also introduced different pricing levels for their models, with higher-cost systems aimed at more demanding uses and lower-cost versions designed for broader access.
The Grok 4.5 launch comes during a period of rapid competition among major artificial intelligence developers. OpenAI is also preparing to release GPT 5.6, which it has described as its strongest model yet. The planned release follows previous limits placed on the model by the Trump administration due to concerns about security implications.
With Grok 4.5, SpaceXAI is positioning its latest model as a faster and more affordable option for businesses and users seeking AI systems capable of handling advanced knowledge-based tasks.
