ChatGPT, Bing, and Bard aren’t the only AI-powered chatbots that can help with research and writing tasks. Claude is one of the services that fits right into this list. However, the question of how well it can handle the same tasks compared to the good old ChatGPT is the one we yet need to answer.
✅ AI Essay Writer ✅ AI Detector ✅ Plagchecker ✅ Paraphraser
✅ Summarizer ✅ Citation Generator
Claude is an AI assistant designed by Anthropic that was first launched in May of 2023. It then upgraded its model from 1.3 to 2.0 in July of the same year. Undoubtedly inspired by the launch of ChatGPT, this bot has some resemblance with this recognized platform but still offers quite a unique website design. However, are there any differences between the two AI platforms aside from their facade?
We already knew what ChatGPT was capable of, so we were mainly curious about Claude’s abilities. After all, we wouldn’t mind switching out the GPT-4.0 subscription to something better, if there was such an option.
ChatGPT vs Claude: What’s the Key Difference?
The main difference we noticed between ChatGPT and Claude was that even though working with Claude was much more pleasant, ChatGPT provided more accurate research results, especially in the case of online sources. To be fair though, Claude does give a warning that the link it provides may not be valid, since the platform is yet on the stage of improvement.
On the other hand, we didn’t enjoy reading the text from ChatGPT, as much as from Claude. The second one did a better job of writing a more human-resembling essay than the first bot.
Despite these differences, our evaluations clearly showed that both platforms provide similar writing quality and capabilities. So no matter which one you decide to use after reading this review, you will still need to double-check the answers from both assistants.
| ChatGPT | Claude | |
| 💼Company | OpenAI | Anthropic |
| 🤖Model | GPT-4.0 | Claude 2 |
| 📧Sign-in | Needed With email | NeededWith email or Google account |
| 💰Free/Paid Use | Free use of GPT-3.5 GPT-4.0 – $20/month | Free use of the beta versionPro version for $20/month |
| ✍️Writing Competence | A generic essay written by a bot, with slightly better references | An essay more pleasant for reading but rather redundant and repetitive |
| 🎯Best fit for | Finding and analyzing relevant sources | Logical structuring of your writing and including in-text citations |
| 🌟A*Help Score | 62/100 | 61/100 |
Claude vs ChatGPT: First View
| ChatGPT | Claude |
| Fully trained language model | The beta version of Claude 2 |
As we noted Claude and ChatGPT resembled each other a lot. Both platforms look like chats where you type in your query and after 30 seconds or so they give you a full answer.
On both platforms, you can ask for references for your research, but ChatGPT is more successful in finding active links with relevant information. This is mostly due to the platform’s recent update which allows the AI to find sources with the help of Microsoft’s Search Engine Bing. Claude, on the contrary, frequently states that it either can’t scour the internet or gives out a warning that it can provide incorrect information, including reference links.
With ChatGPT as well as with Claude you will be able to use the free version of the website or pay $20/month to access a a Pro account. Here, however, lies the major difference. With the first AI, a paid subscription will provide you access to an upgraded GPT-4.0 version. When you make the same purchase with Claude though, you will be working with the same model just with an extended usage capacity.
By the way, we have also earlier compared GPT-3.5 to GPT 4. You can look at what differences we found here.
How We Did the Testing
We try to ensure the highest level of objectivity when conducting any tests with different platforms. To make that happen in the case of comparing ChatGPT and Claude, we used both of them to complete the same task with a particular set of requirements. You can see all the details in the picture below.
We also paid attention to the general experience when using the websites, their functionality, and helpfulness. All of this, alongside writing performance, became the basis for the services’ general A*Help scores.
ChatGPT vs Claude: Results
So, we did our thing, and feed the instructions to the both AI platforms. And here’s what we found.
Generation flow: ChatGPT was Better at Following Instructions
Yes, it’s true that AI is wired to handle detailed prompts. Yet still, ChatGPT was more efficient in staying consistent with our requirements. Claude in this case, decided to improvise, and included not 2 but 3 sources in the reference list.
It’s important to note, however, that Claude offered quite a range of text manipulations. For example, it was more successful in helping us correct grammatical and spelling mistakes in sentences, than ChatGPT was.
Referencing: Neither ChatGPt nor Claude did Proper Referencing
| ChatGPT | Claude | |
| Number of references included | 2 | 3 |
| Compliance with Citation Style | Faintly resembles APA | APA but with too many errors |
Even though one would think that since AI assistants have access to external sources, they would be great at adding and editing references according to publicly accessible manuals. Well, in our case, this assumption wasn’t confirmed by either Claude or ChatGPT. Both services, even despite trying to format references in a specific way, didn’t reach the needed result. Thus our writing was left with a list of courses that faintly resembled APA style.
Wordcount: ChatGPT Proved to be Better at Writing Longer Texts
| ChatGPT | Claude |
| 606 words | 581 words |
When you are given a writing task, most of the time you are also provided with a specified word count for the work. Still, it’s better when you write more. It seems that ChatGPT recognizes that. It wrote quite a long text and covered all the different aspects of the topic. Longer, of course, doesn’t necessarily mean better, but we will see what’s up with that in the next paragraph.
Professor’s Evaluations
| ChatGPT | Claude |
| 30.9/100 | 29.8/100 |
The main idea of our small experiment was to see which AI writing bot – seasoned ChatGPT or fresh Claude – would do a better job of writing an essay. However, to once again stay objective, we decided to go as far as sending both texts for an evaluation by the US professor. As both works were checked according to the generally accepted grading standards, we could clearly see which platform was more suitable for academic writing purposes.
As you can see from the provided grading, Claude did score 1.1 pts less than ChatGPT. In terms of overall writing quality, to be honest, it wasn’t much worse.
Claude won this battle by providing a text with better spelling and punctuation. It was also more logically structured, leading to the text’s higher clarity and reasoning. (If you ask us, this essay sounded more like written by a first-year university student.)
ChatGPT’s text got lower points across all those aspects but was much more grammatically accurate and had a slightly better vocabulary.
As was mentioned before, both services were lost in the citation formatting battle. Nonetheless, ChatGPT’s efforts were marked slightly higher for reference styling, while Claude received a solid “B” for properly including in-text citations. All the same, the quality of cited sources was lacking in both essays.
| ChatGpt’s Essay | Claude’s Essay | |
| Length | 100% | 97% |
| Spelling | 90% | 100% |
| Grammar | 66% | 60% |
| Punctuation | 93% | 97% |
| Word Choice | 74% | 72% |
| Citation formatting | 30% | 23% |
| Efficiency | 6% | 6% |
| Acuity | 6% | 41% |
| Clarity | 31% | 35% |
| Objectivity | 33% | 34% |
| Plagiarism Check | 25.74% plagiarism | 0% plagiarism |
Finally, moving on to the plagiarism levels (because this is the most relevant issue when it comes to AI-generated texts). Below you will see that Claude was a definite winner here as its essay was completely free of any duplicated content. This can be explained by the properly added in-text citations. Hence, ChatGPT’s work, with its 25.74% detected plagiarism, couldn’t stand a chance before Claude’s writing.
Essay by ChatGPT
Essay by Claude
Plagiarism check of the essay by ChatGPT
Plagiarism check of the essay by Claude
The Verdict
If you are on the lookout for a new AI-powered chatbot to use instead of ChatGPT, Claude is hardly a better option. It’s more pleasant to the eye, we can give it that. However, in terms of creating quality content, doing research, and helping with reference formatting it’s not more helpful than the good old service created by Open AI. And even though, de jure, ChatGPT earned by 1 point more on the A*Help scale, in reality, both services provided relevantly the same level of writing assistance.
We didn’t narrow our research down just to Claude though. We also checked out a few other largely famous AI-powered chat bots, like Bard and Bing. Check out those articles to see what we found.
FAQ
Follow us on Reddit for more insights and updates.
More from Academic Writing Tips
Comments (0)
Welcome to A*Help comments!
We’re all about debate and discussion at A*Help.
We value the diverse opinions of users, so you may find points of view that you don’t agree with. And that’s cool. However, there are certain things we’re not OK with: attempts to manipulate our data in any way, for example, or the posting of discriminative, offensive, hateful, or disparaging material.