4 ms·
Been very skeptical for awhile but decided to try it cause it was like $30 for 3 months or something. I cancelled my Claude subscription (still have codex), but
by Jonovono 1mo ago
Been very skeptical for awhile but decided to try it cause it was like $30 for 3 months or something. I cancelled my Claude subscription (still have codex), but been pretty impressed with the grok CLI. Nice TUI and seems intelligent thus far.
- slowin 1mo agoReally? I've used all of the models extensively and Grok doesn't even compare to the others. I can ask for a big task and it will say "Done!" like 3 minutes in, but it will have done just the least amount of work possible. I also notice "muskisms" leaking back in the results like "I'm not roasting your code". Completely unusable for me and not comparable to the other model providers.
- Jonovono 1mo agoInteresting, havn't noticed that. Mine will work for hours. Maybe i'm still in the "get them hooked" phase. Just started using it last week.
- r_lee 1mo agoI'm considering switching as well, used the API to code with Grok via pi, was really pleasant whereas using Claude just gives me a cortisol spike
- slowin 1mo agoThat is interesting, I've never been able to get it to successfully complete big tasks to any level of quality. For me the Anthropic models are by far the best at that, but the newer OpenAI models are starting to approach that quality. Grok always seemed like it was designed to write nasty tweets and the code quality I got from it matched that.
- dotancohen 1mo agoThe only model I've found to beat it for coding (Python, Rust, Kotlin, PHP) and planning is Fable. What are you using it for?
- slowin 1mo agoI'm using it for coding and it's never approached the level of Opus for me. Sol is pretty good too.
- dotancohen 1mo agoI actually haven't used Sol yet. I'd love to hear what you use it for, and how it performs.
- slowin 1mo agoI think it's good for mid-sized tasks. It seems quick too. It just doesn't do as good of a job with large changes or architectural decisions as Opus imo.
- bluecalm 1mo agoWhen did you try it? Grok before 4.5 was way worse in my experience.
- slowin 1mo agoI don't remember the version exactly but maybe June/July timeframe? I tried it on maybe a half dozen projects with different types of asks as I was comparing models for my company at the time.
- rayiner 1mo agoGrok 4.5 came out in mid July and is a huge leap forward. Not quite as good as Opus 5 for code in my experience, but better and more succinct for understanding human documents.
- claaams 1mo ago[flagged]