More in Trip planning & itineraries.
Compare Disney World trip budgets from three AIs with ChatGPT, Gemini and Perplexity
Beginner-friendlyNo logins, payments or coding needed. Low risk: paste the prompt and you should get something useful.
Copy the prompt below and paste it into the AI tool named on this page.
Prompt Original prompt
Exact prompt as published in the source.
Plan a five-day Walt Disney World vacation for my family of five, including two adults and three kids ages 12, 10 and 6. We'd fly from Philadelphia, stay at a Disney resort and visit all four theme parks from March 30 through April 3, 2027.
Try this in:
Sends the Original prompt shown above. ChatGPT and Grok open with it filled in, and ChatGPT may send it straight away. Gemini opens empty, so we copy the prompt for you to paste. Every button also copies it.
Opens in a new tab. We don't see what you send.
These links are only a shortcut. We're not affiliated with OpenAI, Google or xAI, and your results may not match the source.
Optional stronger version
Our rewrite with extra checks. The poster didn't test it.
Before you start, ask me 1–3 short questions about the details below and wait for my answers. Once I've answered, confirm the plan in one line and run the whole task start to finish. Goal: Plan a theme park family trip with a realistic budget for my exact dates. Questions to ask me first: - Who's going, and how old are the kids? - Where are you flying from, and what dates are you considering? - What matters most: lowest cost, fewest lines or the best hotel? Inputs (fill in or confirm): - Travelers and ages: [TRAVELERS] - Home airport and dates: [AIRPORT / DATES] - Parks and hotel type wanted: [PARKS / HOTEL] - Budget ceiling: [BUDGET] Rules: - Check whether my dates fall in a holiday or school-break period and price for that. - Use the park's own age rules for ticket prices. - Give a low and high cost range for flights, hotel, tickets, food and add-ons like line-skip passes, and say which numbers are guesses. - Link the official pages for ticket and hotel prices so I can check them. Checkpoint: Show the cost breakdown before writing the day-by-day plan, and wait for me to adjust it. If you get stuck: If you can't find current prices for my dates, say so instead of guessing low. Verify before you finish: Check the total adds up and every child's ticket is priced at the right age. Output: A cost table, a day-by-day park plan and a short list of what to book first. Done when: I have a realistic budget I've checked against official prices and a plan the whole family can follow.
Customize the stronger version
Answer these questions and we'll put your answers into the stronger version. The original prompt above stays exactly as published.
e.g. 50 dollars
Your prompt
Before you start, ask me 1–3 short questions about the details below and wait for my answers. Once I've answered, confirm the plan in one line and run the whole task start to finish. Goal: Plan a theme park family trip with a realistic budget for my exact dates. Questions to ask me first: - Who's going, and how old are the kids? - Where are you flying from, and what dates are you considering? - What matters most: lowest cost, fewest lines or the best hotel? Inputs (fill in or confirm): - Travelers and ages: [TRAVELERS] - Home airport and dates: [AIRPORT / DATES] - Parks and hotel type wanted: [PARKS / HOTEL] - Budget ceiling: [BUDGET] Rules: - Check whether my dates fall in a holiday or school-break period and price for that. - Use the park's own age rules for ticket prices. - Give a low and high cost range for flights, hotel, tickets, food and add-ons like line-skip passes, and say which numbers are guesses. - Link the official pages for ticket and hotel prices so I can check them. Checkpoint: Show the cost breakdown before writing the day-by-day plan, and wait for me to adjust it. If you get stuck: If you can't find current prices for my dates, say so instead of guessing low. Verify before you finish: Check the total adds up and every child's ticket is priced at the right age. Output: A cost table, a day-by-day park plan and a short list of what to book first. Done when: I have a realistic budget I've checked against official prices and a plan the whole family can follow.
Scroll to see the whole prompt. Your answers show in green.
4 blanks are still empty: Travelers and ages; Home airport and dates; Parks and hotel type wanted; Budget ceiling. They'll stay in brackets in your prompt, so the chat tool may ask you about them.
What you type stays in this browser tab. We don't send it or save it, and it's gone when you leave the page.
AI tools
Outcome
Amanda Caswell gave ChatGPT, Gemini and Perplexity the same prompt: plan a five-day Walt Disney World trip for a family of five flying from Philadelphia, from March 30 to April 3, 2027. ChatGPT came in cheapest, suggesting about $8,500, but its hotel estimate was far too optimistic for those dates. Gemini estimated $10,050 to $13,350 after noticing the dates fall just before Easter Sunday 2027, in a busy spring-break period, and that kids 10 and older pay adult ticket prices. Perplexity's full estimate reached $10,545 to $13,955, more than $5,000 above ChatGPT's figure, and Caswell liked its itinerary best. Caswell judged that Gemini wins on budgeting. When shown the other two estimates, ChatGPT revised its target to about $10,000, with $11,000 available.
Evidence: Article or interviewA newsletter, journalist or interview account of what happened.
Source
- Posted
- by Amanda Caswell (Tom's Guide)
- Last checked
- Method:
- How we verify
Context: All three plans are estimates; Caswell recommends checking prices yourself before booking.
More in Trip planning & itineraries
Plan three family days in Paris, with ticket links, using Copilot
Paolo Asnaghi planned three August days in Paris for two adults and an 11-year-old, using only Copilot (with GPT-5). Asnaghi asked what to see from the hotel's area, excluding Disneyland, then asked for links to buy tickets in advance next to each description. Copilot grouped sights to avoid crossing the city, showed approximate prices and suggested a 3-day transit pass instead of single tickets. In about half an hour Asnaghi had a customized guide. The PDF it made wasn't user-friendly, though. During the trip, Copilot helped find places to eat, explained sights from photos and checked the weather for the next day's plans.
TravelTrip planningCopilotWorkedImproved promptEvidence: Described by posterBeginner-friendlyView CasePlan a five-country family trip with Gemini, despite some wrong tips (partial)
Manish Kurse planned a family trip through France, Germany, Luxembourg, Belgium and the Netherlands with little kids and several families, testing Gemini and ChatGPT along the way. A detailed prompt in Gemini Canvas, covering needs like vegetarian food and kids in strollers, produced a day-by-day itinerary that saved hours and was updated as plans changed. On the road, the AI helped with a lost parking ticket at an Amsterdam P+R, Brussels metro tickets and questions at the Louvre. But a highly recommended Paris creperie was closed, a pizza tip 'didn't sell pizza', and a chairlift in Cochem was shut. Google Maps mostly had it right; 'even Gemini gave the wrong answers.'
TravelTrip planningGeminiChatGPTPartialRecreated promptEvidence: Described by posterView CaseBook a trip, flight, hotel and shuttle with ChatGPT Dots, and keep confirming the same booking (partial)
neom spent two hours trying ChatGPT's Dots agent and grades it B-. Asked to figure out a trip, it did; neom booked the flight and hotel, and Dots noticed the shuttle hadn't been booked and offered to book it with the details they'd discussed. But after neom said yes, it went back to the site, re-checked the details, asked for confirmation, filled in the booking form and then asked again whether to book. It can sign in with passwords, but wouldn't fetch two-factor codes from email even with email access, and its permissions checker rejected broad rules neom wrote, such as proceeding with routine, reversible actions without asking. neom finds it faster to use regular chat.
TravelTrip planningChatGPT DotsPartialRecreated promptEvidence: Described by posterView CaseGet real flight and hotel prices from a Claude Code travel skill, which only guessed price ranges until it was given a browser (partial)
Christopher Johnson built a Claude Code skill in about twenty minutes to act as a travel agent, sorting flights and places to stay for a business trip. Asked for real prices, it gave ranges, such as 'Flights typically run $280 to $420', and none of them was a price they could actually pay. Claude explained that the booking sites load prices with JavaScript after the page renders, which its tools couldn't see, so it had been estimating. With the Claude in Chrome extension installed and the same skill and request, it opened the sites and returned three specific places to stay with current prices and direct links. It was slower and used more of the plan's usage limits. They kept 'Act without asking' off, and Claude never bought anything.
TravelTrip planningClaude CodeClaude in ChromePartialRecreated promptEvidence: Described by posterView CasePlan a sunset hike from a ChatGPT itinerary and get stuck at the summit after the last ropeway had gone (failure)
Dana Yao and their spouse used ChatGPT to plan a romantic hike up Mount Misen on Japan's Itsukushima island. They set off at 15:00 to reach the summit for sunset, exactly as ChatGPT had told them. ChatGPT said the last ropeway down was at 17:30, but when they got to the station to descend, the ropeway had already closed, and they were stuck at the top of the mountain.
TravelTrip planningChatGPTFailureRecreated promptEvidence: Article or interviewView CaseUnderpack for a Mount Shasta climb after following Gemini's advice, ending in a rescue
According to the U.S. Forest Service, three novice hikers planned a Mount Shasta climb partly with Google Gemini, which recommended supplies for an 8-hour climb rather than the 16 hours it took and told them to eat simple carbohydrates rather than fats. They reached the summit at 7 p.m., got lost off-route on the way down after their phone battery died, and one hiker injured a knee at about 11,500 feet. Rangers reached them the next morning; the hikers said they 'relied too much on AI rather than our own critical thinking.'
TravelTrip planningGeminiFailureRecreated promptEvidence: Article or interviewView Case
Related cases
Plan a five-country family trip with Gemini, despite some wrong tips (partial)
Manish Kurse planned a family trip through France, Germany, Luxembourg, Belgium and the Netherlands with little kids and several families, testing Gemini and ChatGPT along the way. A detailed prompt in Gemini Canvas, covering needs like vegetarian food and kids in strollers, produced a day-by-day itinerary that saved hours and was updated as plans changed. On the road, the AI helped with a lost parking ticket at an Amsterdam P+R, Brussels metro tickets and questions at the Louvre. But a highly recommended Paris creperie was closed, a pizza tip 'didn't sell pizza', and a chairlift in Cochem was shut. Google Maps mostly had it right; 'even Gemini gave the wrong answers.'
TravelTrip planningGeminiChatGPTPartialRecreated promptEvidence: Described by posterView CasePlan a Hong Kong trip with Gemini, Grok and ChatGPT (some addresses were made up)
A traveller planned a Hong Kong trip mainly with Gemini, using Grok and ChatGPT as extras, to organise locations, build itineraries, make packing lists and turn store addresses into Google Maps links. The traveller says routes were logical and personalised, but the AIs sometimes produced convincing addresses that didn't exist, and got names, opening hours or whether shops were still open wrong. Next time they'd verify every important detail on maps and official sites.
TravelTrip planningGeminiGrokChatGPTPartialRecreated promptEvidence: Described by posterView CasePlan a 4-day family road trip to South Dakota with five chatbots and use the best plan
A Tom's Guide writer asked five chatbots to plan a four-day family trip from near Minneapolis to the Black Hills and Badlands, with gas and food stops, meals, lodging and under 1,400 miles in a borrowed minivan. Gemini was best at routing; ChatGPT did worst, with a broken map link and a needless night in the Badlands; Copilot was minimal; Claude was good apart from the route. DeepSeek won by basing the family in Custer State Park and planning every meal, and they followed its plan exactly, calling it a blast.
TravelRoad tripsDeepSeekGeminiChatGPTClaudeCopilotWorkedRecreated promptEvidence: Article or interviewView Case
