xAI shipped Grok 4.5 on Wednesday, positioning the model as a cheaper alternative to Anthropic's flagship Opus with comparable performance.

The release marks the first major model launch since xAI went public several weeks ago. The company claims Grok 4.5 delivers "twice greater token efficiency" than competing models while handling coding, research, writing, and office automation tasks.

xAI priced the model at $2 per million input tokens and $6 per million output tokens. That undercuts Opus 4.7's $15 input and $75 output pricing by substantial margins, according to the company's blog post.

Benchmark results published Wednesday show Grok 4.5 performing competitively across standard AI evaluation metrics, though falling slightly short of best-in-class scores in most categories.

"Based on strong positive feedback from customers in our beta test program, @SpaceXAI will make Grok 4.5 available to the public tomorrow. It is an Opus-class model, but faster, more token-efficient and lower cost," Musk wrote on X.

In a follow-up post, Musk claimed xAI's internal assessment shows "Grok 4.5 is roughly comparable to Opus 4.7, but much faster."

Why cost efficiency matters

Token pricing has emerged as a critical battleground as enterprises deploy AI at scale. xAI's efficiency claims, if validated in production use, could give the company significant leverage against OpenAI and Anthropic in enterprise sales cycles.

The model targets the same automation use cases that have driven adoption of GPT-4 and Claude Opus: code generation, document analysis, research synthesis, and complex reasoning tasks.

xAI plans to make Grok 4.5 publicly available Thursday, expanding beyond the beta customer base that has been testing the model in recent weeks.