Chatgpt 5 Pro vs Grok 4

[White Night]

I just upgraded to the ChatGPT Pro plan to try out GPT-5 Pro, and honestly, it surprised me quite a bit. But it’s pricey, and the usage limits on the Pro version make me feel it’s not really worth the money. What’s been your experience?

I’m also thinking about switching to Super Grok instead—has anyone here tried Grok and compared it with GPT Pro (especially Grok 4)?

 

[4440]

To be honest, i use many models concurrently. And i mix use them by a lot, like feeding response from LLM A to LLM B.

This depends on your use case, you should see which one gives better results for your specific use case.

If you don't want to waste money, buy some low credits API-keys and test them on a free webgui, then make decesions.

My view is I uses both gpt-5 and grok-4, and i know when to use gpt-5 and when to switch to grok-4, for my own use cases.

 

[Ranking Expert]

If i need to choose 1 then i go for GPT

 

[White Night]

I’ve noticed a problem with ChatGPT 5 Pro:
When I tell it something incorrect, or point out that it didn’t mention a certain detail, it usually just adapts to my statement and says, “Sorry, you’re right.”

This makes me feel like it’s not being truthful, but instead simply changing its answer based on whatever I say.

 

[Promogram]

GPT is much more capable and does pictures pretty well unlike grok.

 

[SirTop]

To be honest, i use many models concurrently. And i mix use them by a lot, like feeding response from LLM A to LLM B.

This depends on your use case, you should see which one gives better results for your specific use case.

If you don't want to waste money, buy some low credits API-keys and test them on a free webgui, then make decesions.

My view is I uses both gpt-5 and grok-4, and i know when to use gpt-5 and when to switch to grok-4, for my own use cases.

This is my experience too with both after trying them out. There are always cases where one does something better than the other... just like with any other LLM.

I mainly still use ChatGPT because I have a lot of ongoing projects there and it helps me stay organized with so many things.
But if their prices keep increasing the way they are I might switch to something less expensive : )

 

[White Night]

Recently, I became interested in Thai amulets after watching a series on Netflix. I’d like to learn more about them and find information that categorizes the different types of amulets by their purpose or intended wishes. However, I’ve noticed that even with ChatGPT Pro, the recommendations aren’t entirely accurate or complete. The responses often miss important details in the research, and I only get more input when I ask very specific questions about why.

wonder if grok 4 can do a better job

 

[tubeboost]

ChatGPT is a champion at the moment, maybe because they launched its model first.

 

[Fiducia]

Grok is just unpredictable, political and inconsistent. You can't do any automation with it nor use it for business.

OpenAI, Mistral and eventually Deepseek are way better.

 

[seoways]

I would choose ChatGPT without any hesitation.. My favorite so far. Grok is better too, but still cant beat ChatGPT.

 

[jansha]

Personally GPT still gives better answers. Haven't tested the new chinese one yet.

 

[vickysan]

Claude is best so far for my use, I use it for coding and it's just better. I tried grok's coding model and it just cannot keep code context.

 

[Venusaur]

( ) grok
( ) chat gpt
(x) gemini

 

[Panther28]

Grok does video now! from a image, so i can get a high quality image from chatgpt, trace it in Animate, make it super high quality, then drop it in grok, and say make char walk, or run, or do whatever, and i get 2 video creation options! Who needs veo!? lol

 

[ActualOutreach]

The best thing to do is to use BOTH, not just one of them. To compare each time each one of their answers and to compare.

 

[White Night]

just try grok 4 beta yesterday, and chatgpt is quite better in my opinion
but in my point of view, the version 4o is quite better & accurate than version 5 or version 5 thinking

 

[Proxy Cake]

To be honest, i use many models concurrently. And i mix use them by a lot, like feeding response from LLM A to LLM B.

This depends on your use case, you should see which one gives better results for your specific use case.

If you don't want to waste money, buy some low credits API-keys and test them on a free webgui, then make decesions.

My view is I uses both gpt-5 and grok-4, and i know when to use gpt-5 and when to switch to grok-4, for my own use cases.

With GUI do you recommend?

 

[andrew mapple]

I just upgraded to the ChatGPT Pro plan to try out GPT-5 Pro, and honestly, it surprised me quite a bit. But it’s pricey, and the usage limits on the Pro version make me feel it’s not really worth the money. What’s been your experience?

I’m also thinking about switching to Super Grok instead—has anyone here tried Grok and compared it with GPT Pro (especially Grok 4)?

I agree with you in my opinion grok is more usefull then ChatGPT

 

[andrew mapple]

Grok is too good but according to my chatGPT it is also good but late process in making any kind of Images

If you want to generate images, use Gemini. Its AI image generator is better than ChatGPT's. I personally use it many times when ChatGPT doesn't make the image I want.

 

[4440]

With GUI do you recommend?

i use my own but any gui is fine. there is no significant difference. for ease of use, you can try Cherry Studio, it is portable, no install needed, great for casual chat or quick testing.