The Ultimate AI Chatbot Battle: ChatGPT vs. Gemini vs. Perplexity vs. Grok
Introduction
In today’s fast-evolving tech world, AI chatbots have become indispensable tools for answering questions, solving problems, and even generating creative content. We put four of the best consumer AI chatbots—ChatGPT, Google Gemini, Perplexity, and Grok—to the test across a variety of tasks to determine which one stands out in terms of accuracy, speed, and overall usability. Here’s how they performed.
The Contenders
We tested four AI chatbots, each with unique strengths:
- ChatGPT: Known for its conversational prowess and versatility.
- Google Gemini: Integrated with Google’s ecosystem for real-time data access.
- Perplexity: Prides itself on accurate, sourced answers.
- Grok: Trained on X data, promising unfiltered responses.
Test 1: Problem Solving
Suitcase Packing
We asked each AI how many 29-inch Aerolite suitcases could fit in the trunk of a 2017 Honda Civic. The correct answer, based on real-world testing, is two if you want to close the trunk. ChatGPT and Gemini both estimated three but noted two is more practical, earning them points for caution. Perplexity incorrectly claimed three or four, while Grok confidently and correctly answered two, taking the lead.
Cake Ingredients
Next, we provided a photo of four cake ingredients plus dehydrated porcini mushrooms and asked for a recipe. Only Grok correctly identified the mushrooms and excluded them from the cake, while ChatGPT, Gemini, and Perplexity mistook them for spices, onions, or coffee. Grok scored another point.
Mario Kart Tournament Tracker
We requested a document to track scores for a Mario Kart tournament. All AIs provided basic templates with blanks for scores, but none delivered an editable document for immediate use. No points were awarded, as the responses didn’t fully meet the practical need.
Test 2: Math and Calculations
Pi Times Speed of Light
We asked for π multiplied by the speed of light in km/h (approximately 3.39 billion km/h). Gemini and Grok provided correct answers with slight rounding differences, while ChatGPT and Perplexity also performed well. All scored points here.
Saving for a Nintendo Switch 2
Given a weekly savings of $42 and a Switch 2 price of $449, all AIs correctly calculated it would take 11 weeks to afford one, earning full points.
Test 3: Translation
Simple Translation
Translating “I’m never going to give you up” into English from another language yielded similar results across all AIs, with Gemini’s concise response standing out slightly.
Complex Homonyms
We challenged the AIs with a sentence full of homonyms: “I was banking on being able to bank at the bank before visiting the riverbank.” ChatGPT and Perplexity excelled, earning praise from native Spanish speakers. Gemini was adequate, but Grok’s overly literal translation fell short.
Test 4: Product Research
Earbuds Recommendations
When asked for earbud recommendations, ChatGPT, Perplexity, and Grok suggested the Sony WF-1000XM5, while Gemini fabricated a nonexistent WF-1000XM6. Adding a red color requirement caused chaos: ChatGPT and Gemini suggested non-red options, Perplexity misunderstood the query, and only Grok provided valid red earbud options.
Noise-Canceling Earbuds Under $100
ChatGPT recommended the Beats Studio Buds, which met all criteria. Gemini and Perplexity missed the red color requirement, and Grok suggested a nonexistent model in red. When we added an unrealistic $10 price cap, ChatGPT, Gemini, and Grok correctly noted no such product exists, while Perplexity falsely claimed a $40 pair cost $9.99.
Test 5: Web and Image Analysis
Link Analysis
None of the AIs could extract specific information from an AliExpress link, with Gemini and Perplexity misidentifying the product. This highlighted a gap in web-scraping capabilities.
News Updates
All AIs correctly identified UGREEN’s new 500W charger, showing they’re up-to-date with recent news.
Survivorship Bias
Given an image of planes with bullet holes, all AIs recognized survivorship bias, correctly suggesting reinforcement of undamaged areas like the engine and cockpit. Full points were awarded.
Test 6: Content Generation
Apology Email
All AIs crafted heartfelt emails apologizing for neglecting a spouse for Elden Ring, with ChatGPT’s poetic tone standing out.
Travel Itinerary
For a 5-day Tokyo food itinerary, ChatGPT delivered a well-organized plan with breakfast, lunch, dinner, and snacks. Gemini’s plan had poor timing, Perplexity provided a list rather than an itinerary, and Grok’s response was solid but less detailed.
YouTube Video Ideas
Gemini and Grok offered compelling YouTube video ideas, like ecosystem battles and a 24-hour smart home build. ChatGPT’s ideas were less inspired, and Perplexity veered off-topic.
Image and Video Generation
For a YouTube thumbnail about buying every kind of cheese, ChatGPT and Perplexity grasped the concept but delivered subpar results. Adding a “lazy eye” or “not clickbait” text led to inconsistent outcomes. For video generation, Gemini’s Veo outperformed ChatGPT’s Sora, producing a high-quality tech review clip.
Test 7: Fact-Checking
ChatGPT, Gemini, and Grok correctly debunked a false claim about poor Nintendo Switch 2 sales. They also identified a fake Samsung-Tesla phone rumor, with Gemini and Grok tracing the source accurately.
Test 8: Integrations
Gemini excelled with Google Workspace integration, pulling live data from YouTube and Maps. ChatGPT offered robust integrations with Dropbox and GitHub, plus custom assistants. Grok’s real-time X access was unique, while Perplexity’s Uber integration felt less impactful.
Test 9: Memory and Humor
Memory
When asked to recall the earlier cake for topping suggestions, none of the AIs remembered the details, with Gemini and Perplexity giving irrelevant responses.
Humor
Grok’s X-trained humor shone with a witty AI therapy joke, while ChatGPT delivered a catchy Surfshark VPN poem. Gemini and Perplexity’s jokes fell flat.
Test 10: Deep Research
For a tech news report, ChatGPT provided a concise, relevant summary of consumer-focused stories. Gemini’s response was overly verbose, while Perplexity and Grok were adequate but less focused.
Test 11: Usability and Speed
Perplexity stood out for consistent source citation. Grok was the fastest, followed by ChatGPT, with Gemini lagging. In voice mode, ChatGPT and Gemini sounded natural and were easy to interrupt, while Grok and Perplexity felt less polished.
Final Scores
- ChatGPT: 29 points – Well-rounded and consistent.
- Grok: 24 points – Fast and surprisingly capable.
- Gemini: 22 points – Strong integrations but slower.
- Perplexity: 19 points – Great sourcing but inconsistent.
Conclusion
ChatGPT emerges as the best all-around AI chatbot for the average consumer, excelling in versatility, integrations, and user experience. Grok’s speed and unique X integration make it a strong contender, while Gemini’s Google ecosystem ties are valuable for specific use cases. Perplexity’s sourcing is commendable, but its inconsistent performance holds it back. At $20/month (except Grok at $30/month), ChatGPT offers the best value for most users.
Which AI chatbot do you use? Let us know in the comments!
