This reads like a sponsored article promoting xAI, without clear ethical disclosure. While it appears to be about AI-generated games in general, it focuses solely on Grok-generated content.
I use it as my primary coding assistant, when I'm able to. I haven't paid for the more advanced models from others, and it seems to be the most advanced free-to-use thinking model at the moment.
Aider can't use grok 3 with thinking yet, afaik, because xai hasn't made it available in the API.
From what I'm hearing, it and Claude 3.7 "thinking" are very similar in performance.
I've spent a lot of hours vibe coding with sonnet 3.7 thinking and I'm not seeing anything in the article that jumps out at me as being different from my experience.
Other models often perform better (https://web.lmarena.ai/, https://aider.chat/docs/leaderboards/) - I have yet to meet anyone who uses Grok as their primary programming assistant.