Six travel prompts have been circulating, all built on the same idea: hand ChatGPT or Claude the role of a fare expert, and it hands back the deal the booking sites hid from you. Half of them do real work. The other half ask a language model for live pricing data it does not have, and it will answer anyway, in full sentences, with numbers.
Worth knowing which is which before you plan a holiday around one.
1. The route decoder
You are an expert flight hacker. I want to fly from [city] to [destination] around [dates] with a budget of [amount]. Give me five routing strategies the booking sites don’t surface: which nearby airports to check, which days shift the price by hundreds, which layover combinations come out cheaper, and which booking window to aim for. Be ruthlessly specific.
This one holds up, because most of what it needs is geography and airline structure rather than today’s fares. A model knows Charleroi is Brussels for Ryanair purposes and that Hahn is nowhere near Frankfurt. Treat the output as a shortlist of searches to run, not as a price quote.
2. The price drop predictor
You are an airfare analyst. Here is the route I’m watching: [describe]. Tell me when to buy based on this route’s historical pricing patterns, which days of the week prices drop, and whether waiting or booking now is smarter. Then tell me what to set alerts for and where.
Here is where it breaks. No model has your route’s price history, and the general rule keeps moving: Expedia’s air travel report named Sunday the cheapest day to book in 2025, then Friday in 2026. A “best day” that changes every year is not a pattern. It is noise with a press release attached.
Keep the last sentence of the prompt. Alerts on Google Flights and Skyscanner are real and free. Delete the rest.
3. The hidden city audit
You are a fare structure expert. Analyse this route: [describe]. Find every unconventional booking strategy that cuts the price: hidden city tickets, positioning flights, open jaw routings. Explain the risk of each so I can decide what I’m comfortable with.
I have dropped “mistake fare windows” from the original, because mistake fares are accidents. Nobody has a window on them.
The rest is worth asking about, as long as you read the word “legal” carefully. Hidden city ticketing is not a crime. It is a breach of the contract you agreed to at checkout. Lufthansa took a passenger to court over a skipped final leg, and the case ran through the German system for years. Airlines can void your return, close your frequent flyer account and bill you the difference. Open-jaw routings, by contrast, are entirely ordinary and the airline will happily sell you one. Ask the model to keep those two categories apart.
4. The accommodation arbitrage
You are a travel cost strategist. I’m visiting [destination] for [X] nights with a budget of [amount]. Find me the neighbourhoods with the best value, what to search on each platform, which dates to shift if flexibility saves real money, and what I should never pay full price for in this city.
Fine, and the neighbourhood question is the strong half. Which district sits twenty minutes further out at half the price is stable, well documented and exactly the sort of thing you would otherwise get from a friend who lives there.
5. The total cost reality check
You are a blunt travel budget advisor. Here’s my trip as planned: [describe]. Find every cost I haven’t accounted for: baggage fees, tourist taxes, airport transfers, tipping culture, dynamic pricing traps, and the three things that always catch out first-time visitors. Give me the real number.
The best of the six, and the least exciting. Tourist taxes and baggage rules sit in published documents and change slowly. This is the prompt that saves you real money, mostly by pointing out that Amsterdam charges a percentage of your room rate and that Ryanair’s cabin bag isn’t what you are picturing.
6. The rebooking radar
You are a flight monitoring strategist. I booked [route] at [price]. Tell me how to watch for price drops that trigger this airline’s rebooking policy, whether it offers free changes, and what to say to customer service to get a credit if the fare falls after I book.
Useful in the United States, largely decorative in Europe. The 24-hour free cancellation rule that most of this advice rests on comes from the US Department of Transportation, and it does not apply to a KLM ticket. European carriers set their own terms and mostly don’t refund the difference. Ask about your specific airline rather than about airlines in general, and check the answer on the airline’s own site before you phone anyone.
What actually transfers
The pattern across all six is the same: these prompts work when they ask about structure and fail when they ask about price. Airport geography, fare rules, tourist taxes, baggage policy, and neighbourhood value are written down somewhere and hold still long enough for a model to learn them. Today’s fare on your route is not, and no amount of role-play in the prompt gives the model eyes on it.
Use them as research, then go and check. The model narrows the search. It does not do the booking.
