On Wednesday (local time), Elon Musk's technology enterprise, SpaceXAI, officially rolled out its latest large-scale model, Grok 4.5. This launch marks the first significant product release since the company's public debut a few weeks prior. According to a blog post from SpaceXAI, Grok 4.5 is earmarked as a 'flagship model' tailored for general-purpose applications. It is engineered to manage a diverse array of tasks that the AI sector is currently striving to automate, encompassing code writing, application development, office and clerical duties, academic research, content creation, and other repetitive knowledge-intensive tasks.

SpaceXAI asserts that Grok 4.5 boasts a Token utilization efficiency that is double that of comparable mainstream models. Should this efficiency be confirmed through real-world application, it could substantially cut down on usage costs. Official benchmark test data indicates that Grok 4.5 is on a par with top-tier competitors in several capabilities, although it has not yet attained comprehensive superiority across all dimensions.

Moreover, Grok 4.5 has broadened its scope of applications, rendering it apt for more specialized fields such as legal and financial services. Simultaneously, it has bolstered its cybersecurity features. The model underwent joint training by SpaceXAI and AI programming startup Cursor, which was valued at $60 billion in the relevant transaction.

Regarding pricing, Grok 4.5 charges $2 per million input Tokens and $6 per million output Tokens, costing less than half of what comparable models demand. Presently, Grok 4.5 is accessible through the SpaceXAI console, GrokBuild, and Cursor. Users can commence utilizing it free of charge via a web link.