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.

Sorry, english is not my first language. Itss disabled on our work API instances so my use is strictly personal because it was available on Cursor (my account is pre-acquisition). I have extensive AI experience across nearly all models, open weight and frontier and closed, I am very aware of how they talk or output. This is high/xhigh, via Cursor, not via X Grok/UltraGrok if that make different.

Grok 4.6/4.7 feel like you are less tokens because Cursor gives you double the quota subsidising Grok specifically - there is a separate Grok/Cursor model quota bar in addition to non . Its actual token drain if you are paying API rates is generally far higher than Fable, Astra et al. You are not using less tokens.

It is weird in some reward way, it will frequently, at least for me, complete the task in the laziest way possible,Technically it is done, but that's about it.

For example a user registration system ended up with users being able to log in as anyone because the authn was a cookie set with the user's pkey id, unsigned. Nearly anything it outputs technically works but if you give it a GLM or DS pass you will find dozens to hundreds of vulnerabilities of varying hilarity (Fable refused, Opus refused).

I have seen it write SQLI-vulnerable code, asked it to review a file without saying what's wrong and it did not catch it (fresh session, single file @-tagging in harness). Again, technically, the code works and does what the prompt asked, its just handing in some of the laziest copied homework I have ever seen.

If you specifically call out to use prepared statements it will typically put a plaster over this however I have also seen Grok 4.6 use prepared statements by concenating the user input raw into a statement then executing it with no ? or named replacements, rendering it somehow a SQLI-vulnerable prepared statement. One med it added a OR 1=?, replaced the ? with another 1, then said OK you are using prepared statements now.

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.