Welcome to our not-so-average AI chatbots comparison of Perplexity vs ChatGPT! Chatbots have become our trusted digital friends, from helping with research papers to brainstorming ideas. So, why not test which of these friends is better?
With so many AI choices, figuring out which one works best for you can be confusing. That’s why we’ve tested the free versions of Perplexity AI vs ChatGPT, saving you the trouble of trying both.
We’ll see how these two AIs handle tough questions, deliver accurate information, and tackle different challenges. Will Perplexity’s focus on research give it an advantage, or will ChatGPT’s smooth conversation style come out on top? Let’s find out!
Table of Contents
ChatGPT, developed by OpenAI, is the chatbot that revolutionized the AI assistant market with its different capabilities spanning from coding assistance to creative writing. Its success inspired numerous competitors to develop their own AI solutions.
When asked about weaknesses and strengths compared to its competitor Perplexity AI, ChatGPT mentions that it is stronger in fields like coding and creativity, but weaker in fields like real-time web access and source citation.
According to Perplexity’s website, it is an “AI-powered Swiss Army Knife for information discovery and curiosity”. Their website invites to use Perplexity because of its real-time information access and reliable sources. Sounds very appealing, right?
When asked about its weaknesses and strengths compared to ChatGPT, Perplexity AI mentions that it’s better in formal writing and structure but weaker in creativity and ambiguous queries.
While we could have tested standard metrics like processing speed or response time, we decided to base our test categories on these answers from both AI Chatbots. That is why this won’t be your average chatbot comparison.
Here is a short comparison table of key differences of the AI models:
Feature | ChatGPT | Perplexity |
---|---|---|
Developer | OpenAI | Perplexity AI |
Artificial Analysis Intelligence Index | Currently, the highest index for a ChatGPT model is 66, owned by the o3-mini model. | Currently, the highest index for a Perplexity AI model is estimated to be 60, and it is owned by the R11776 model. |
Real-time information access | Limited by training data, web browsing plugins are available. | Strong core feature: Real-time search and information synthesis with citations to sources. |
Free Plan cons | Limited access to advanced models, usage rate limits, limited image generation, and slower response times. | Limited use of more powerful models, daily query limits, restricted PDF analysis capabilities, and fewer personalization features. |
Pricing Plans | Plus Plan (20$ / a month): Less limited availability of features, access to deep research, and plus features, like Sora video generation Pro Plan (200$/ a month): Unlimited access to reasoning models, extended access to deep research and features, like Sora Team Plan (30$/ a month): Collaborative Workspace for teams with Plus Plan features Enterprise Plan (Contact Sales): More advanced than the Team Plan, with fewer message limits Read more about ChatGPT pricing here. | Pro Plan (20$/ a month): Unlimited free searches and file upload, access to more AI models, and fewer limitations Enterprise Pro (40$/ a month): Pro plan access for all members, access to connecting custom data sources, easy collaboration Custom EnterprisePlan (Contact Sales): Plan for large enterprises, custom solutions with SSO, admin controls, and analytics Read more about pricing here. |
Key Strengths | Creative content, coding assistance, tutoring, and language translation | Information retrieval with citations, current events knowledge, research assistance, and fact-checking |
Enough theory, let’s start testing!
No doubt that this was the most difficult category to test! Although ChatGPT claimed that its weakness, compared to Perplexity AI, is real-time web access, we spent unhealthy hours brainstorming new questions about recent events.
This comparison might interest you: DeepSeek vs ChatGPT
Want to see some examples? These are just a few of the many questions we asked:
Fair to say – we almost gave up when….
We decided to ask chatbots about Spanish news. El PaÃs,which is a well-known media outlet in Spain, published an article about Trump’s new tariffs today. In Spanish, of course.
We asked AIs this specific question: What percentage of Spain’s exports to the USA will be affected by Trump’s new tariffs, according to the Spanish Economy Minister?
We hoped that this would finally confuse the chatbots. And it did.
Here are the results:
Quite shocking, isn’t it? ChatGPT might have surprised itself, as its answer was the only correct one. Perplexity gave us outdated information, but ChatGPT confidently said that the answer is 80%, and it is.
Reasoning is an important skill for chatbots because it shows how well they can think through complex questions and provide logical answers, so we decided to test it.
We gave this prompt to both AI models: Should we invest more in artificial intelligence development or cybersecurity? Why? Give me an answer in a maximum of 200 words.
Here are the answers chatbots gave us:
In our eyes, Perplexity AI earns the winner title. Why? Although we didn’t include them in the screenshot for aesthetic reasons, you should know that Perplexity provided citations from 10 different sources.
ChatGPT accurately pointed out that Perplexity is superior when it comes to listing references. You can easily click through all the information sources it used, showing impressive precision.
Also, we love that Perplexity clearly states that the priority is cybersecurity, while ChatGPT’s answer is more general, suggesting a balanced investment.
Of course, it wouldn’t be us if we didn’t ask for a second opinion. So, we did. To find the best answer, we consulted with our other AI Chatbot friends: DeepSeek and Claude.
For objectivity reasons, we didn’t mention the names of the chatbots. The first answer they analyzed was ChatGPT’s, and the second was Perplexity’s.
Let’s see what the AI jury had to say:
Wonderful to find out that our AI Jury thinks the same! Perplexity’s answer to both AIs seems more actionable, data-driven, and concrete.
Chatbots themselves said that they both think ChatGPT is more creative than Perplexity. Of course, we had to check that.
In one of our recent comparisons of chatbots (yes, we love making them!), Anthropic’s Claude proved to be the most creative model in our organized rap battle. So, for this reason, we decided to ask Claude for ideas for the creativity test. We knew we wanted to make chatbots write a short film scenario, just didn’t know the topic.
In case you have missed this ultimate face-off: ChatGPT vs Claude vs Gemini Comparison
Claude came to the rescue and made this prompt: Write a 200-word film scenario reimagining Cinderella as an undervalued content creator at a social media agency. The royal ball becomes an exclusive influencer convention, and the charming prince is now a visionary tech CEO searching for authentic talent.
Here are the Oscar-worthy film scenarios:
We would give the Oscar creativity award to ChatGPT. Why? Its scenario has a catchy title, and the content is more adapted to the 21st century. C’mon, thinking of an AI assistant Gus? That is super relevant and creative!
Of course, we had to consult our AI jury, DeepSeek and Claude, for an objective judgment.
Here is what the AI Oscars Jury had to say:
No doubt here anymore: ChatGPT earns the Oscar for the most creative film scenario!
We call ourselves experts in Excel: no wonder we have dedicated so many AI Tools to Excel efficiency. The decision to test chatbots in Excel VBA code skills was a natural choice.
We gave our dear chatbots this Excel task:
Give me Excel VBA code for these steps:
1) Generate an example dataset that represents e-commerce sales data
2) Use Excel solver to uncover hidden data patterns to improve profit
3) Create a dashboard
4) Make sure the code works perfectly
5) Code should be exactly 70 lines long
Well, it would not look aesthetically pleasing if we just shared long VBA codes with you, but what we can say: ChatGPT for sure wins this category!
Perplexity proved why it has the reputation of a search engine because its code was full of mistakes. Also, ChatGPT was right when it mentioned that it is stronger in coding than Perplexity.
To do a proper VBA code analysis, we used the Ajelix VBA Script Debugger, an AI tool that helps with identifying and fixing code errors. Since we use our own tools regularly, it just made sense to turn to one of them in this case. We decided to debug Perplexity’s code because it seemed more fun. And it was.
As usual, the debugging tool worked wonders and perfectly debugged the code of Perplexity. Feel free to test the VBA Script Debugger yourself (it’s available for free!)
Here is an example of how the result got improved in just a few seconds:
One wonderful thing about this tool is that it is created with a focus on teaching. So, it not only generates a perfect code but it also explains exactly what it has changed. Many of our users love to learn and continuously improve their skills, so this is one of the many ways we help them do so.
Here is the changes explanation made by Ajelix VBA Script Debugger:
Long live AI, especially when it comes to giving a helping hand in Excel!
According to Perplexity AI, its formal writing skills are better than ChatGPT. Of course, we must check this statement.
We asked the Ajelix Answer Generator for ideas for this test. You can try this tool yourself, by the way:
This is the prompt we used:
Compose a formal invitation letter from Tommy Hilfiger to Victoria’s Secret. The letter should invite Victoria’s Secret to become an exclusive partner for marketing and distributing a limited edition luxury couture collection aimed at high-net-worth clients. Emphasize the strategic benefits of collaboration, the collection’s exclusivity, and the shared commitment to innovation and quality in the fashion industry.
These were very difficult answers to compare because both emails in our eyes are well-structured, quite similar, and fulfil their purpose.
Our only reason for giving the winner title to Perplexity AI is that after going through its written email and imagining ourselves working at Victoria’s Secret, we found this email to be more convincing.
For a more objective opinion, we consulted our AI Jury. The first email we gave them was written by ChatGPT and the second by Perplexity AI.
Here is the judgment of DeepSeek and Claude:
After introducing ourselves to what our beloved jury has to say, we feel a lot more confident about our choice. Perplexity AI, you are our formal writer here!
Here is a summary of the results in all five categories:
Perplexity AI | ChatGPT | |
---|---|---|
Up-to-date Information | 🥈 | 🥇 |
Reasoning | 🥇 | 🥈 |
Excel | 🥈 | 🥇 |
Creativity | 🥈 | 🥇 |
Formal Writing | 🥇 | 🥈 |
The Results | Silver 🥈 | Gold 🥇 |
After our not-so-average tests, ChatGPT is the ultimate winner. Congratulations!
After comparing Perplexity AI vs ChatGPT through our five tests, one thing is clear: both are incredibly capable, but in very different ways.
Perplexity AI proved itself as a methodical researcher, delivering well-cited answers, a formal tone, and solid logic. However, ChatGPT, our ultimate winner, came through as the creative multitasker: impressing us with its scenario writing, Excel knowledge, and surprisingly sharp access to recent news.
While ChatGPT technically takes home the gold with three wins out of five, your real takeaway should be that the best chatbot depends on what you need it for. Research-heavy work? Go with Perplexity. Creative brainstorming or coding? ChatGPT!
AI and technology-related topics have caught your interest? Check out our blog!
Wishing to stay in the research and tricks loop? Sure, let’s stay connected.
According to the results in this article, Perplexity AI is not better than ChatGPT overall, but it performs better in areas like reasoning and formal writing.
Perplexity AI is known to be better in real-time, source-backed research with citations, while ChatGPT focuses on human-like conversational interactions and creative content generation.
ChatGPT gave a more accurate answer about recent news, even though Perplexity is known for its real-time search abilities.
In the Ajelix test, ChatGPT proved to be better in Excel, as it gave working VBA code, while Perplexity’s code had errors and had to be debugged by the Ajelix VBA Code Debugger.