Scoop: SpaceXAI launches new model, Grok 4.5
TL;DR
SpaceXAI is launching Grok 4.5 as its first model since going public and acquiring Cursor. The release is aimed at coding, agentic work and knowledge work, not mainly consumer chatbot use. The company says Grok 4.5 was trained alongside Cursor and beats comparable models on engineering and knowledge-work benchmarks. So far, the evidence is mostly company-supplied charts.
Nauti's Take
Treat the Opus-class claim as vendor framing until independent tests cover real coding tasks, tool use, and cost per completed task. For small teams, the first useful check is whether Grok 4.5 inside Cursor is actually more stable or cheaper than the current Claude, Codex, or Gemini setup, and what data leaves your environment.
Briefingshow
Grok 4.5 matters less as a chatbot update and more as another push toward AI models embedded directly inside developer tools and agent workflows. If the speed and pricing claims hold up, the comparison shifts from raw intelligence to cost per completed task. The catch: the strongest claims come from SpaceXAI itself, and independent tests are still missing.
Sources