01My expectation going in was low
I expected Grok to feel like a PR play — something built to say "we have an AI too" rather than something genuinely useful. After using it for a few weeks, I have a more complicated view.
It has some real strengths that no other AI model currently offers. It also has clear weaknesses that make it a poor choice as your primary tool. Here's what actually came from testing it.
02Three things Grok does that others don't
Real-time X/Twitter data is the most significant one. Grok has access to what's being posted on X right now. Ask it what people are saying about a topic today and it actually knows. Claude and ChatGPT don't have this.
The restrictions are noticeably looser. Grok will engage with questions that Claude tends to deflect or water down. Whether that's useful depends entirely on what you're using it for.
Built-in image generation through Aurora is included for X Premium subscribers. Quality is decent for quick visuals.
03Where it's weaker
General reasoning, complex coding problems, detailed analysis — Claude and GPT-4 are meaningfully better. Grok is improving but it's not there yet for serious technical work.
It's also deeply tied to the X ecosystem. If you're not already using X regularly, the real-time social data feature doesn't matter much to you.
04Free access
Basic Grok is free on X without a subscription. The more capable Grok model requires X Premium. The free version is usable but limited.
For most people, Grok makes most sense as a specific-use tool alongside your main AI — particularly when you need current information from social media or recent news — rather than as a ChatGPT replacement.
05Bottom line
Better than I expected. Not as capable as the best models for general work. If you need real-time social media intelligence baked into your AI queries, Grok is the only option doing that right now.
Worth trying if you have an X account. It takes two minutes and the free tier exists.
