ChatGPT Dots vs Grok Bot: what I learned using them side by side

For a few weeks now I've been using ChatGPT Dots and Grok Bot side by side and, for now, they're neck and neck. Both have things I like and things that frustrate me, but neither has an advantage the other couldn't pick up within two weeks. There's one scene that sums up pretty well what I've been noticing: I asked Grok Bot to send an email, it sent it and marked the task as done. The email bounced and it never found out. I only found out because my ChatGPT dot told me on its own.

That difference interests me because it changes quite a bit how I evaluate these assistants. How much autonomy to give them, how to share context, and whether it's enough for them to just do what you asked. I don't think there's a clear winner yet, but after using them for several weeks I have five fairly concrete impressions.

Several bots or just one?

Grok Bot's ability to create several bots and have them talk to each other seems spectacular to me. You can split responsibilities and conversations instead of putting everything in one place. Dots, for now, limits you to just one, so topics end up getting pretty mixed together. Grok Bot has a big edge there. I prefer being able to keep topics separate, although having a single dot with all the context has its advantages too.

Which one is less restrictive?

Grok Bot is way less restrictive. You ask Dots to log into a bank and it asks you for authorization for everything. Grok Bot doesn't seem to care much about anything, hahaha. In practice you can feel that difference, because with one you can move more freely and with the other you have to step in a lot more.

I don't necessarily think that granting fewer permissions is always the right call. There are actions where I'd rather the agent check with me, especially if they can have consequences that are hard to undo. But it's also true that if every step needs approval, at some point delegating stops saving you much work.

Which one is more proactive?

Here I lean toward Dots. It doesn't need me to set up a routine to follow up on the topics we've been talking about, while with Grok Bot I'm not seeing that initiative as much. The difference shows up especially when a task has already been executed but there's still something left to sort out. The bounced email was a pretty clear example of that.

There's a huge difference between completing a task and understanding when the result you were after still hasn't been achieved. Grok Bot sent the email, but didn't notice it never arrived. Dots, without me asking, told me about the problem. I don't know if they'll always behave like this, but so far that proactivity makes me trust Dots a bit more for things I want followed up on.

Which one handles hard problems better?

When I give them something that requires thinking a bit deeper, I feel Grok Bot struggles more. It's not that it can't solve complex tasks; my impression is that Dots finds a better way to approach them. In those situations it opens a separate conversation with ChatGPT, like the one we'd normally use, and works the problem from there.

I like that option because not everything you ask an agent needs the same treatment. Some things are straightforward; others need you to stop and think before acting. In my experience, Dots handles that second category better. I didn't run a controlled evaluation of the two, and I'm not trying to present this as a definitive difference in capability: it's what I've been noticing in everyday use.

What about memory and context?

This point probably explains a good part of my current preference. I've been using ChatGPT for years and Dots takes advantage of all the context it already has about me. It knows previous conversations, topics I kept coming back to and things I'd otherwise have to explain again. With Grok Bot I'm still building that base, so the comparison doesn't start from exactly the same place.

And transferring context takes time. One explanation isn't enough, because there are always conversations and references that Dots already knows and Grok Bot doesn't yet. Part of the value is in what the agents can do, and another part is in how much I need to explain for them to understand what I want.

Accumulated context is starting to be a huge advantage, even when the other system may be just as capable or more. That's also why I don't know how much of my leaning toward Dots comes from the product and how much from the years I've spent using ChatGPT. I think it's important to acknowledge that before drawing conclusions that are too firm.

ChatGPT Dots vs Grok Bot: comparison as of October 2026

These are some of the differences between the products as of October 2026. The table helps you get your bearings, but it doesn't replace what happened to me using them.

So, which one should you use?

If I had to stick with one today, I'd pick Dots, but by a very small margin. It feels more proactive to me, I feel it handles problems that need more analysis better, and it has the huge advantage of having known me for years. Grok Bot, on the other hand, I love for the ability to work with several agents and for how few restrictions it puts in place. I wouldn't rule out my answer being different a few weeks from now.

The interesting thing is that I care less and less about which one answers an isolated question better and more about how it works when I delegate something to it. Do I have to stay on top of it? Can it pick a topic back up on its own? Does it notice when something went wrong?

For now, I lean just slightly toward Dots, with the suspicion that a good part of that edge may come from context and not necessarily from it being better. Because if I have to review every step and be the one telling it something went wrong, it's not taking much work off my plate.

What are you using? Have you started delegating tasks to them, or do you still mostly use them to chat?

Updated: October 2026