ChatGPT vs. Perplexity: How writers use them and which AI sidekick to choose
- Hafsat Ahmed
- Sep 28
- 11 min read
Confession: I once swore I'd never use AI, telling myself it was about being principled.
But if I'm being honest, I was scared too.
When ChatGPT launched in 2022, the writing team I was part of dissolved. Around the same time, Harvard Business Review reported a 30.37% drop in writing jobs.
Between that and the flood of online “hot takes”, steering clear of AI felt safest. So much so that I turned “I don't use AI” into a pitching strategy.
Today, my perspective has changed. While I don't let AI write entire drafts, I've integrated it into my workflow – mainly for research.
And I'm not alone.
A recent LinkedIn poll I ran showed that most writers now incorporate AI into their workflows.
Clearly, in 2025, avoiding AI doesn’t give you an edge. It might just make you slower and less competitive.
So, where should you start if you're getting started with AI?
In dozens of conversations with writers, two tools kept coming up:
ChatGPT – creative, flexible, and conversational.
Perplexity – fast, precise, and citation-backed.
Both are best-in-class but serve different purposes. I tested them across writer-focused tasks and spoke with writers who use them.
This article will cover what each tool does best, where it falls short, and when to use which one.
Quick overview: ChatGPT vs. Perplexity at a glance
Here's a quick snapshot:
Features | ChatGPT | Perplexity |
Primary strength | Conversational, flexible, creative content generation | Research, speed, real-time answers with citations |
Best use cases for writers | Examples, interview prep, gap analysis | Summaries, YouTube transcribes, gathering research resources |
Response style | Direct, skimmable, concise | |
Web access | Yes, via SearchGPT | Outpaces ChatGPT, designed primarily to access live data |
Similarities | Web search, text generation, voice/image features, project/space organisation, deep research, file uploads, analysis | |
Key differences | Runs on OpenAI’s GPT models, conversation history retention, chat-like interface, creative ideation, longer responses | Supports multiple LLMs, independent queries, citation-backed answers, research-first focus, real-time sourcing, concise output |
Pricing |
|
|
How writers use ChatGPT
ChatGPT significantly propelled public interaction with AI. Since entering the scene, it has consistently given competitors a run for their money.
Its chat-style interface feels almost human, which is why it has become a staple in my workflow, as well as that of many writers. In fact, every writer I have interviewed so far uses ChatGPT in some way.
Here's how they make it work:
1. Finding super-specific examples
Nothing sells your arguments like a solid example, but sometimes finding them takes forever.
Before ChatGPT, freelance content writer Aleksandra Beka Jovičić could spend an hour searching for worthwhile examples. “Now I find super specific examples in minutes,” she says.
The best part? You'll often find examples you wouldn't discover the old-fashioned way.
Try Aleksandra's prompt:
“Find me an example of (specific subject) that (specific action or characteristic) in (timeframe), including (details you need) and (context or outcomes).”
2. Spotting blind spots in your arguments
However stellar your writing is, there's always room for improvement. Sometimes you're too close to see the gaps.
Freelance content writer Rosanna Campbell uses ChatGPT to gain that outside perspective. “One of my favourite use cases for ChatGPT is having it analyse my arguments. I'll use a prompt like:
‘Here's a summary of my thesis. Pretend you're [industry expert] and pull it apart. Please show me where I'm wrong, where I'm oversimplifying, and what I'm missing. Be ruthless.’”
This helps her identify weaknesses in her reasoning and strengthen her drafts.
Note: ChatGPT sometimes invents issues simply to give feedback. Know when to stop and trust your judgment on what's worth addressing.
3. Double-checking content briefs for alignment
Before submitting drafts, I give them one final review. I upload the content brief alongside my draft and ask ChatGPT to flag any misalignments.
I'll use a prompt like:
“Attached is my draft and the content brief. Please compare the two and highlight:
Any missing sections or requirements
Areas that don’t align with the brief’s instructions
Suggestions to better match the brief”
This quick pass cuts my editing time and ensures I've fully met the brief.
4. Staying consistent with client preferences
Being a good writer is one thing; being easy to work with is another. The first gets you clients, the second keeps them.
One way to be easy to work with is by maintaining consistency with client preferences, such as intro length or style. ChatGPT can track these patterns.
Take SaaS content writer Oluwaseun Akinlembola, for example. He uses ChatGPT to keep recurring issues in check: “I upload my feedback doc and ask it to edit my work. I accept what I agree with and discard what doesn't feel right.”
5. Simplifying complex terms
If you're treading unfamiliar ground, ChatGPT can smooth bumps by breaking concepts into plain, digestible language.
SaaS content writer Funmito Obafemi, who writes for a relatively new industry, takes advantage of this:
“ChatGPT helps me break these topics into simpler terms to create engaging content for my audience.”
Try it using this prompt:
“Explain (technical term) in simple, non-technical language for a specific audience (e.g., beginner marketers). Give me a short analogy and an example of how it works in practice.”
6. Switching voices for global audiences
I once had to write for UK and US audiences in the same week.
Since I was more familiar with writing in US English, I spent hours adjusting phrasing and hunting for regional differences to make the UK piece a fit.
With ChatGPT, it's simpler. Upload your draft and ask it to highlight how to “UK-ify” (or “US-ify”).
Here's a prompt you can use:
“Review this draft and adapt it for the (UK/US) audience. Highlight spelling, idioms, and examples to localise the tone and phrasing.”
Resource: Quality prompts guarantee better outputs. For a quick head start, try Ruben Hassid's ChatGPT prompt maker.
Now let's see how Perplexity compares.
How writers use Perplexity
I first tried Perplexity when a fellow writer shared it to help me find the source of an interesting statistic I'd been hunting. It surfaced the original within seconds.
That's how many writers see it: an AI search assistant. It provides real-time, cited answers in a straightforward style – distinct from ChatGPT's conversational tone.
But it does more than fact-checking. Here's how it fits into writing workflows:
1. Finding relevant research resources
Perplexity is especially handy for gathering credible research resources in one go.
Use a prompt like:
“Find relevant case studies, whitepapers, reports, and statistics about AI in content marketing.”
Tip: Specify a timeline (ideally 2-3 years) so your references stay current.
2. Transcribing and summarising YouTube videos
When I'm pressed for time or want to confirm a video's content before watching, Perplexity gives me the gist.
I'll often use a prompt like:
“ Act like a professional transcription assistant. TRANSCRIBE the attached YouTube link or CLEAN the raw transcript provided. Deliver a polished transcript with timestamps every 20–30 seconds or when the speaker changes, so it's easy to follow along.”
Tip: You can also specify that it removes filler words, unless contextually relevant.
3. Digging into Reddit conversations
Reddit is one place to find honest, unfiltered discussions from users – but it can be a time sink.
Perplexity speeds this up by pulling relevant threads in one go.
“Perplexity is my go-to for digging into Reddit conversations,” says Oluwaseun. “I understand Reddit has Answer AI now, but I prefer getting the summary in one place, and Perplexity does that.”
Here's a prompt to try out:
“Find active Reddit threads from the last 12 months where people discuss (SPECIFIC TOPIC) in detail.”
4. Summarising long reports
If you have dense research reports to analyse with looming deadlines, use Perplexity to summarise them into clear key points.
Try it with my prompt:
“Summarise the linked report about (TOPIC). Focus on the key findings, major comparisons, important statistics, and main conclusions. Keep the tone neutral and concise, and structure the summary into three paragraphs.”
5. Verifying competitor features
When writing competitor reviews, the last thing you want is to spread incorrect information.
Instead of scanning multiple pages, ask Perplexity to confirm features and cite its sources.
SaaS marketer Sreyashi Chatterjee shared this effective prompt:
“I am writing a review article on [Tool name] and want to know whether the [Tool] offers this particular feature [Feature name]. Please verify and share the relevant source.”
Tip: It's also handy for finding differentiating features in comparison pieces: Try “What unique features does (product A) have over (product B)?”
6. Tracking trends in your industry
If you write thought-leadership content or work in fast-moving fields like AI, Perplexity can help you track emerging trends across platforms like Reddit, X, and LinkedIn.
Prompt it like this:
“What are the top emerging trends in (Industry name) being discussed on Reddit, X, LinkedIn, and review sites in the past 3 months. Include sources.”
This pulls together fresh insights and keeps you up to date.
ChatGPT Vs. Perplexity: head-to-head comparison
I tested both on five writer-focused tasks using identical prompts. To keep things fair, I chose tasks that rely on features they share.
For each task, I evaluated accuracy, speed, usability, and depth of response.
Disclaimer: These tests were run on ChatGPT Plus (GPT-5o) and Perplexity Pro in September 2025; your experience may vary with subsequent updates.
1. Conducting deep research on a niche topic
I find it invaluable that these tools enable quick exploration of multiple dimensions of a topic.
Right off the bat, I was excited to see how their deep research feature would handle the “implications of AI adoption in SaaS content marketing.”
ChatGPT first asked clarifying questions, which added to its collaborative feel. It returned 27 sharp sources in about eight minutes – that was worth the wait. Its narrative was straightforward, quoting industry reports and citing data.
What impressed me most was how it coherently connected the findings, rather than just listing them. It stayed strictly on topic and leaned more towards industry-wide resources, such as the Deloitte 2024 year-end GenAI report, which was exceptionally loaded.
On the contrary, Perplexity took under 3 minutes and surfaced 82 sources. Some were tangential, but it also pulled strong case studies and reports. Now I noticed a difference in its approach. Its style was direct and data-driven, citing metrics such as a “20-30% higher ROI” from McKinsey.
What stood out for me was the vast amount of recent resources Perplexity referenced. ChatGPT also included newer sources, but Perplexity accessed real-time 2025 data, like the SaaS Capital survey.
Both performed impressively, each winning where the other falls short. ChatGPT takes the crown for analytic depth and strategic organisation. Perplexity wins for its speed and fresher sources. So, it's a split.
You can access both reports here
Winner: Tie
2. Analysing & summarising a lengthy report
Next, I asked both to summarise McKinsey's 2025 State of AI report in under 300 words.
ChatGPT responded in under a minute with a digestible summary, including headers for easy reference. It delivered on the prompt I gave, but leaned toward a broader approach.
On the other hand, Perplexity took about two minutes and returned a consumable overview – I got the gist in a skim. Its approach was granular, providing precise specifics from the report.
So who won? Both nailed the task, but slightly overshot the word limit. ChatGPT gave me a concrete background context. However, Perplexity's summary provided specific data points and findings that I could drop straight into my writing.
Winner: Perplexity
3. Turning raw data into a visual layout
Both support AI image generation, which I've recently been experimenting with. ChatGPT uses DALL-E, while Perplexity restricts it to pro users.
So, I asked them to visualise data from McKinsey's 2025 AI report.
ChatGPT's design checked most boxes – clean layout, accurate data representation, and even included the source. The formatting was polished, with well-labelled axes and annotations that indicated percentage changes.
Meanwhile, Perplexity's output was a sleek, minimalist design with contrasting colours that were eye-friendly. The layout felt uncluttered, but it omitted contexts I requested, such as source references.
While Perplexity's design was sleeker, ChatGPT's inclusion of source and detailed annotations made it more practical.
Winner: ChatGPT
4. Real-time web search & information retrieval
This is one of Perplexity's strong suits, but ChatGPT (using SearchGPT) seems to be catching up.
So I asked both to cover the OpenAI job opening controversy in 200 words.
ChatGPT's real-time web search impressed me. In seconds, it delivered every detail in a skimmable format – job role, salary range ($310k–$393k), location, and more, with citations.
I loved how it used quotes to capture the reaction of the content marketing community to the news.
Perplexity delivered solid results with 12 source links, including the official job postings and news outlets. The factual accuracy was spot-on, and it captured similar community sentiments.
Both exceeded the word limit, but nailed the task. I'll pick Perplexity if I need extensive source verification for research. However, ChatGPT provided the freshest community reactions and presented the information in a format that I could immediately digest.
Winner: ChatGPT
5. Preparing targeted interview questions for sources
For my final test, I chose interview preparation, as it's a standard part of writing workflows.
Sticking to the theme, I asked both to develop interview questions for an internal expert on “the aftermath of AI adoption in their SaaS content marketing strategy.”
ChatGPT's questions were engaging and well-sequenced. With a few edits, I can see myself using them. The specificity, such as “operational challenges that content ops teams face,” demonstrated knowledge about the interviewee and their role.
Interestingly, Perplexity's output seemed similar to that of ChatGPT. It generated logical questions, but they lacked the specificity necessary to elicit strong answers. For example, it asked, “How has AI transformed workflows…?” while ChatGPT asked, “How has AI adoption shaped the way your team aligned content priorities…?”
The second question is more likely to generate detailed, quotable responses. That pattern held across most of the other questions. For that, ChatGPT takes this one home.
Winner: ChatGPT
TLDR? Here's a tabular summary.
Task | Winner | Why |
1. Deep research | Split | ChatGPT's depth, sharp sources, and organisation. Conversely, Perplexity's speed, comprehensive resources, and real-time data |
2. Summarisation | Perplexity | Perplexity's precise specifics made its summary easier to implement |
3. Visual data layout | ChatGPT | Translated the prompt better, with a balanced aesthetic and data context |
4. Real-time web search | ChatGPT | Structured, skimmable info on the job with the freshest community reactions |
5. Preparing interview questions | ChatGPT | More specific, strategic questions for quotable answers |
Note: If you hire writers, don't worry. Skilled writers know when (and when not) to lean on AI. If you're unsure, ask them to explain their process.
ChatGPT & Perplexity: what to watch out for
As handy as these tools are, they have their quirks, including:
1. Instruction compliance issues
Both occasionally exceeded my word limits. ChatGPT also pulled older materials despite my instructions. So expect to make manual adjustments where needed.
2. Performance varies by task complexity
Complex tasks can be misinterpreted. Instead of feeding them long drafts or transcripts, break requests into 1-2 focused sections with clear questions for better results.
3. Hallucination and falsification happen
Sometimes, ChatGPT confidently generates incorrect information. Perplexity isn't immune either. So never accept anything an AI tool gives you without verification.
4. Usage limitations and access restrictions
The free tiers may suffice if you're starting out and need only occasional brainstorming or research.
However, for regular use, you'll quickly hit a wall. For example, ChatGPT's free tier allows only 5 deep research queries per month, whereas Perplexity allows 3 queries per day (roughly 90 per month).
That means paid plans (both starting at $20/month) might become necessary quickly.
So which one should you choose: ChatGPT or Perplexity?
Having tested both across writer-focused tasks, here's my verdict:
ChatGPT proves highly versatile. It excelled at most tasks, delivering depth, structure, and polish.
Now, Perplexity isn't playing second fiddle either. Though narrower, it holds its own with unbeatable precision, speed, and stellar source verification.
Do you have to choose? Not really, you can use both. Use ChatGPT for interview prep and content analysis, Perplexity for fact-checking and research gathering.
The bottom line is, AI isn't a threat. Used well, these tools amplify your writing career. And as Lizzie Davie put it in Friday Freelance Tips, “the future of freelancing isn't us vs. AI, but us WITH AI.”
So maybe the right question is: which one will you work WITH and HOW?
Better yet, pick any prompt from this article and test it out with ChatGPT or Perplexity now. See which one fits your workflow better!
Comments