SpaceXAI Releases Grok 4.5, Musk Calls It an ‘Opus-Class Model’

SpaceXAI has released Grok 4.5 on Wednesday, marking the company’s first major model launch since its public listing several weeks ago. In a blog post, SpaceXAI positioned the new release as a versatile workhorse capable of handling coding, app building, office and clerical work, research, writing, and other forms of routine knowledge work.

Elon Musk announced on X that SpaceXAI would make Grok 4.5 publicly available on Thursday, citing strong positive feedback from beta test customers. He described the model as an “Opus-class model,” directly comparing it to Anthropic’s premium lineup, and later added that internal assessments show Grok 4.5 is roughly comparable to Opus 4.7 but significantly faster.

Pricing is a major differentiator: Grok 4.5 costs $2 per million input tokens and $6 per million output tokens, compared to Opus 4.7 at $5/$25 and OpenAI’s top-tier Sol at $5/$30. SpaceXAI claims the model offers “twice greater token efficiency” than other leading models, which if validated in real-world usage would represent a dramatic cost advantage for heavy users.

The company released benchmark metrics alongside the announcement, showing Grok 4.5 is competitive with top models from rivals but slightly short of best-in-class performance. This suggests a pragmatic strategy that prioritizes cost-effectiveness over absolute benchmark supremacy.

The launch comes during an exceptionally busy week for AI model releases. OpenAI is planning to release GPT 5.6 (codenamed Sol) on Thursday, which it calls its “strongest model yet,” after its release was previously delayed by security reviews under the Trump administration. The near-simultaneous arrival of two major models signals an intensifying competitive landscape.

For enterprises and developers, Grok 4.5’s pricing could accelerate the ongoing trend of falling inference costs. If its efficiency claims hold up, xAI could carve out significant market share by undercutting OpenAI and Anthropic on price. However, Grok still lags behind both competitors in developer tooling, API ecosystem breadth, and third-party integrations — areas that will determine whether cost alone is enough to win adoption.

The key questions going forward are whether Grok 4.5’s real-world performance matches its benchmarks and whether SpaceXAI can leverage its public market capital to quickly close the ecosystem gap. OpenAI’s Sol release this week will provide the first real test of Grok 4.5’s true competitive positioning.