I bought a month of use for $100. Early impressions of Grok 4.6 was that it's good at talking about ideas (got me unstuck from a piece of writing that I was working on; that bought it a TON of goodwill) and just okay as a coding tool compared to my more extensive use of Fable and Opus. Not as smart as Fable but cheaper; not as capable as Opus 5 but way less annoying along the way. And it generates images, which Anthropic doesn't do.

Not sure if I'll hold the subscription but I could see myself working with it more.

I use it in the car to talk to because it's built-in, mostly for navigation via voice. It seems to be an old or quantised model, because it seems a few years old, and it has strict rate limits unlike Gemini, where it really just shuts off if you talk to it a few times in one day, but it's still fun to show passengers who are still new to LLMs.

I tried it last month after hearing here that it didn’t speak in Claudisms.

As a chatbot it’s totally fine, virtually indistinguishable from Gemini or ChatGPT or Claude.

For coding it’s… okay. I tried 4.6 and it feels similar to Opus from 12 months ago, or maybe Sonnet from 9 months ago. YMMV.

[flagged]

The only people I know of in the UK who will tell you that they use it are performatively alt-right or right wing. The kind of people who say "nanny-state" or "wokerati". GB News viewers. People who have an opinion on Meghan Markle that they think other people need to hear.

It's just an observation but so far a pretty solid correlation. Musk has so severely poisoned the well in terms of his UK reputation that the only people who are open about using Grok are... well, wankers is as good a word as any.

FWIW among the AI-using people, it mostly goes Claude Code, then Codex, then whatever runs on their Mac. The only Cursor user I knew has jumped ship to OpenCode.