Scoop: SpaceXAI launches new model, Grok 4.5
TL;DR
SpaceXAI is launching Grok 4.5, its first model since going public and acquiring Cursor. The pitch is coding, engineering and agentic work, not another consumer chatbot demo. Musk frames Grok 4.5 as an „Opus-class model“: faster, more token-efficient and cheaper than comparable systems. SpaceXAI claims it beats Opus 4.8 on several engineering benchmarks.
Nauti's Take
The announcement is heavy on PR, but the strategic point is real: SpaceXAI is not selling Grok 4.5 as a clever chat toy, but as a work model for code and agents. Cursor matters more here than another benchmark bar chart.
For anyone in Europe, the hype has to wait. The useful test, once available: one real repo, one real bug and a clean cost comparison against Claude, OpenAI and existing Cursor setups.
Briefingshow
Grok 4.5 is less a chatbot story than a move into the developer workspace. With Cursor, SpaceXAI gets distribution inside the IDE, where models touch real files, tests and repositories. For teams, the practical question is no longer just benchmark rank, but whether the agent is reliable enough in daily work and whether cost, permissions and EU access line up.
Sources
