ChatGPT was sort of unarguably the pioneer of LLM’s. For me, at least, it was the worlds introduction into AI back in 2022, and it kept its status for a few years whilst everybody played catch up.

It was really 2025 where competition picked up out of nowhere. Microsoft Copilot, Google Gemini… and Anthropic’s Claude. These models weren’t necessarily released in 2025, but they skyrocketed into popularity within no time.

Now, each LLM is offering different perks and USPs, and I have seen a lot of people cancelling their OpenAI subscriptions and jumping ship – not just to Claude, but mainly to Claude. But why? What makes Claude so different and amazing that people are cancelling their subscriptions to be part of the gang?

Today I’ll compare ChatGPT and Claude based on:

• Knowledge and Intellect
• Tone and Personality
• Capability
• Limitations and Constraints

Knowledge and Intellect

4 volatile years into mainstream AI usage, both models have been heavily trained and therefore are highly knowledgeable. This makes it harder to judge which is truly ‘smarter’, because it’s not like one will be able to answer a question that the other will have no idea about.

If we are talking about an SAT, they’re both going to score the same, but outside of basic Q&A’s, they have different strengths…

ChatGPT is generally used and preferred for brainstorming, quick tasks, and as a glorified/higher-quality search engine for any and all daily questions. Outside of searching, it is commonly used for drafting emails, reports, homework essays and language translation.

On the other hand, Claude excels in deep data analysis, thanks to its larger content window, offering 200,000 to 500,000 tokens (roughly 150,000 to 350,000 words) depending on the tier, compared to ChatGPT’s standard 128,000 to 256,000 tokens.

As I’m about to discuss, Claude is generally more natural and nuanced in it’s writing, making it increasingly favourable for long-form professional writing and proofreading. I’ll be honest – I use Claude to proofread my blogs! (Never write them, though – don’t get it twisted.)

Whether an AI is ‘smarter’ than the rest is largely subjective and inconclusive, but there are definitely strengths and weaknesses of both. In my opinion, the model with deeper analysis skills and better recall would be classed as ‘smarter’, in my books, so I’ll give Claude the first win.

Tone and Personality

One thing I strongly dislike about AI is its isms. ‘You’re not crazy.’ ‘Firstly, breathe’, ‘And honestly, you’re right to think that!’ – it’s horrific and makes me cringe every time. But that isn’t specific to Claude or ChatGPT… not necessarily.

ChatGPT is more widely noted as being overly praising and corporate, kind of like a ‘yes man’. By default, ChatGPT is more enthusiastic and eager to please.

Claude, on the other hand, is more often cited as more thoughtful and measured, who adopts a more natural, and ‘human’ (if you can call it that) tone. But, having used Claude, I have sometimes noticed the same isms that I have encountered across all LLMs previously.

When writing long passages, Claude is often less detectable as overly AI and more natural. However, I do pose that this could be because ChatGPT is more widely used and has been used by many for much longer, so therefore perhaps it is easier for people to pick up on it’s tone, because we have been exposed to so much more of its content.

Now, tone and personality isn’t too much of an issue with LLMs nowadays, as both Claude and ChatGPT allow you to input custom instructions for how you would like your LLM to speak and operate. However, it’s worth noting that when I primarily used ChatGPT, I did input custom instructions, but since using Claude more, I have not inputted custom instructions but have not had any qualms with the way it speaks to me.

Overall, as a baseline, I’d say that Claude wins in terms of tone and personality, that is, if the criteria is sounding most natural and human.

Capability

ChatGPT is commonly known as a Swiss Army Knife, acting as the all-in-one place for language, image and video generation, also having coding and mathematical capabilities.

I remember using DALL-E (OpenAI’s image generation software) before ChatGPT was a thing! So ChatGPT definitely has the upper hand in terms of being experienced and seasoned in what they offer. Not to bring another LLM into the mix, but I have personally had a better image generating experience with ChatGPT than I have with Gemini’s Nano Banana – but that could just be me!

However, whilst ChatGPT does have coding abilities, it’s pretty well known that Claude takes the cake when it comes to coding. Claude Code has absolutely blown up in recent months. In just 13 MONTHS, Claude Code’s usage volume grew by 42,896 times , and now is not only used for reviewing and debugging, but for entire web and app generation (like hey, that’s my job!).

I wrote an entire blog post about Vibe Coding, and Claude Code has a major part in that industry currently, whereas ChatGPT isn’t even in the running. Claude’s coding capabilities also help their chatbot, with some writing prompts resulting in full landing page creations, whereas ChatGPT simply outputs the text.

Both LLMs are so strong in different fields. By miles, Claude takes the cake when it comes to coding, but it cannot natively generate images and videos, something that ChatGPT is very well-suited for.

ChatGPT is still known as the all-in-one, but perhaps not forever! For now, I’ll give them the win on this round.

Limitations and Constraints

Claude has caused more frustration than ChatGPT when it comes to limitations.

ChatGPT Plus (£20/month) and ChatGPT Go (£8/month) allow users to send up to 160 messages with their top model (currently GPT-5.3) every 3 hours.

On the other hand, Claude Pro (£16/month) has a roughly 45-message limit every 5 hours.

This is a HUGE difference. People have expressed a lot of frustration that they’re so limited to their queries, especially when paying $20 a month. ChatGPT, for this reason, is much more accessible – even with their free plan you can send 10 messages with their newest model every 5 hours.

However, I would argue that based upon my previous discussions with what each model is used for, you wouldn’t necessarily be sending as many messages to Claude as you would to ChatGPT, as you would be less likely to be rapid-fire spit-balling ideas with Claude like you might with ChatGPT. Either way, the frustration is rife on Reddit, X and beyond.

Claude has also been said to struggle with ambiguous prompts. It is more formatted to work with lots of detail and input, so is therefore not as useful when you’re in the very early stages of ideation, or don’t have a clear structure in your mind of what you’re trying to get out of your conversation with Claude. This is another reason that ChatGPT is preferred for brainstorming, as it can arguably be a better ‘thinking-buddy’, which you can then develop on with Claude.

On a personal level, I have also struggled with Claude’s voice and multimodal tools. Speaking is a far more efficient and productive way to use AI, but Claude responds slowly and very robotically, which I have found to interrupt the flow of conversation and thought process. On the other hand, ChatGPT is incredible for voice responses. I have also tried Gemini, but neither have compared to ChatGPT. It is natural, perhaps eerily human, and speaks quickly, which I LOVE for when I’m in the zone.

I have also used ChatGPT when speaking and practicing my Spanish, and it speaks at a much more natural speed than other LLMs that speak too slowly and therefore aren’t the best reflection of real life conversations. Maybe not as relevant to all of you, but something to consider if you ever use AI bots for language translation or learning.

However, I’ll cut Claude some slack here and say that they haven’t had their spot in the limelight for as long as ChatGPT, and therefore aren’t as advanced when it comes to updates and new iterations. However, that isn’t too much of an excuse as Claude did only come out 4 months after ChatGPT in March 2023.

For this final round, I will crown ChatGPT as the LLM that has less limitations and constraints.

Overall

It’s a tie! I know, how boring. I suppose it’s difficult to compare Claude and ChatGPT to one another when they are so commonly used for different things. As I mentioned previously, using them in conjunction with each other, perhaps by beginning an idea with ChatGPT and growing it in Claude, is not a bad way to go!

I recently saw a Reddit thread that said ChatGPT and Claude are currently going back and forth every 3 months with winning over the public until on of them either messes something up or does something that makes them slightly better. I’m sure this battle will continue for a while longer, and I’m almost certain that more LLMs will come into the running sometime soon.

What do you think? Are you more of a Claude or ChatGPT user? I’d be interested to hear your thoughts! Email me jamie@withjamie.co.uk or drop a comment on socials!

I guess I should proofread this blog post with Gemini, huh? No biases here in the withjamie headquarters.