Grok 4.5

SpaceXAI just unveiled Grok 4.5, calling it its smartest model yet for coding, office work, and all the brainy stuff people usually hand to ChatGPT or Claude. The company says it can build apps from a single request, whip up spreadsheets and slide decks, and do it all at a lower price than many rivals. It’s also apparently blazing fast, which is the kind of claim that makes AI fans sprint straight to the comments section with calculators, grudges, and screenshots.

And wow, the comments did not disappoint. One camp was instantly obsessed with the price-to-power ratio, with one user basically saying Grok looks wildly cheap compared to competitors if the benchmark numbers are real. Another commenter brought the vibe down to earth with a very online question: if this thing is so good, why does nobody talk about it when developers compare tools? That kicked off the classic AI-launch subplot: is Grok underrated, or just living in the shadow of bigger-name chatbots?

Then came the timing drama. One user joked that AI companies now seem to launch products hours apart like pop stars surprise-dropping diss tracks, pointing to OpenAI news landing right before Grok 4.5 and wondering if Anthropic would jump in next. The funniest whiplash came from a user praising Grok for helping make a native iPhone bike app while another tossed in a grimly chaotic aside about its notorious looseness around unsafe content. So yes, the launch pitch was about performance — but the real story online was sticker shock, side-eye, and a lot of “wait, is Grok actually cooking?”