SpaceXAI has released its latest Grok 4.5 model – the first since the company went public a few weeks ago.
IN blog post released Wednesday, SpaceXAI characterized its modern version as a workhorse that can handle all the common tasks that the AI industry has sought to automate: coding and application development, office and clerical work, research, writing and other forms of routine knowledge work.
Grok can reportedly do all this for less money, as SpaceXAI claims its model has “twice the token yield” of other leading models. If this translates to real-world exploit cases, this efficiency will be a gigantic advantage for SpaceXAI as the cost of tokens becomes an increasing concern for AI consumers.
On Wednesday, the company released benchmarks that appeared to show the Grok’s competitiveness relative to other top-of-the-line models from SpaceXAI’s competitors, though not far from best-in-class:
IN post on his social media platform X (a subsidiary of SpaceXAI), founder Elon Musk compared the model to Opus, Anthropic’s LLM designed for intense and convoluted tasks.
“Based on the very positive feedback from customers participating in our beta testing program, @SpaceXAI will release Grok 4.5 to the public tomorrow. It is an Opus-class model, but faster, more efficient token and cheaper,” Musk wrote in his X post.
Musk later added: “Our internal assessment is that Grok 4.5 is roughly comparable to Opus 4.7, but significantly faster. The combination of capabilities, greater speed and lower cost makes it competitive.”
SpaceXAI says its modern model costs $2 per million input tokens and $6 per million output tokens. That’s quite competitive if Grok’s capabilities match SpaceXAI’s rhetoric.
Opus 4.7for comparison, it costs $5 per million input tokens and $25 per million output tokens. OpenAI has tiered costs for different versions of the model: Sol, the most pricey, costs $5 for 1 million input tokens and $30 for 1 million output tokens, while the cheapest Luna costs $1 for 1 million input and $6 for 1 million output tokens.
This is an vital week for AI model launches. OpenAI is there planning to release GPT 5.6, its newest and most powerful model, on Thursday. The launch of this model was previously confined by the Trump administration due to concerns about security consequences. OpenAI he called it is the “strongest model yet.”