Going somewhere soon? Practice the conversations before you board. Airport check-in, hotel front desk, restaurant ordering, asking directions, handling a problem with your booking. Voice-to-voice roleplay with an AI that plays the staff. Free.
Start practicing - freeNo credit card. Sign in with Google.
Memorizing 'where is the bathroom?' doesn't help when the airline gate agent asks if you're checking a bag and you don't catch the question. Travel mode roleplays the actual exchanges - turn-by-turn, with the natural variations real staff use.
Airport: check-in, security, gate changes, missed connection. Hotel: reservation, late checkout, room problems. Restaurant: menu questions, dietary needs, asking for the bill. Plus: SIM cards, taxis, asking for directions, lost luggage, currency. Cycle through whichever you need this week.
Real airline staff speak fast. Real hotel concierges have accents. The AI partner adjusts pace and register - clearer at beginner, full natural speed at advanced - so by the time you actually travel, the speed isn't a shock.
Travel English is a strange dialect. It's not formal enough to be business English and not casual enough to be conversation. It's transactional - you have a goal, the staff has a goal, you both want to finish quickly and politely. The vocabulary is specific (boarding pass, late checkout, gluten-free, exchange rate) and the phrasing patterns are predictable. Which is good news: it means you can prepare for it.
The bad news is that the standard phrasebook approach doesn't work. You memorize 'I would like to make a reservation' and then the front desk says 'Have a reservation already?' and you're stuck. The phrase you memorized was the question, not the answer. Real travel exchanges are turns of conversation, not single phrases. You need to practice both sides of the dialogue, in real time, with natural variations.
fluentwith's Travel mode runs you through those dialogues turn-by-turn. The AI plays the front-desk concierge, the gate agent, the waiter, the taxi driver. You play yourself. The AI starts the exchange (most travel conversations are staff-initiated), and you respond. When you stall or use the wrong phrasing, the AI doesn't break character - it just moves on, the way a real staff member would, leaving the awkwardness in the conversation. The post-session report tells you what to fix.
Each scenario has the natural follow-ups baked in. Airport check-in isn't just 'I'd like to check in'; it includes the bag question, the seat-preference question, the priority-boarding upsell, and the 'is your name spelled correctly here?' double-check. Hotel reservations include the credit-card hold question, the breakfast-included question, the late-checkout request. The conversations end the way real ones do - with a polite signoff and a small piece of information you might miss if you weren't listening.
Common pitfalls the report catches: over-formal phrasing (saying 'I would be most grateful if you could possibly' when 'could you please' is right), wrong question structure ('how much it costs?' instead of 'how much does it cost?'), filler-heavy nervousness ('um, like, I think maybe...'). Most travelers fix the same five issues - the report gets specific so the fixes are too.
Recommended practice: pick the one scenario you're nervous about for your next trip. Run through it twice in advance. Then before you leave the hotel each morning of the trip, do a 5-minute warmup. Speaking the language out loud the morning of beats studying the night before by a wide margin - fluency is a motor skill, and motor skills are easiest in the hour after you've used them.
Eight core situations cover 90% of trips: airport check-in (and security, gate, missed connection), hotel reservation and check-in, ordering food at restaurants, asking and giving directions, handling a problem (lost luggage, broken AC, wrong charge), getting a taxi or ride, asking about prices and paying, and basic small talk with locals. Travel mode rotates through all of these.
Start with three scenarios: ordering food, hotel check-in, and asking directions. These three account for half of all interactions on a typical trip. Practice each one ten times in different variations. By session 5 the rhythm of the conversation will feel automatic, which means even when the staff says something unexpected you'll have enough fluency to recover.
Three tactics that work: (1) ask them to repeat - 'sorry, could you say that again?' is universally understood and not rude; (2) ask one question at a time even if they ask several - 'first, the seat - yes, by the window, please'; (3) confirm what you heard - 'so the gate change is to B7, correct?'. We practice all three.
More ways to practice English speaking with AI: all modes
Casual conversationInterview EnglishBusiness EnglishVideo practiceNetworking EnglishEnglish debateStorytellingPronunciation
Travel mode uses neutral English by default. The vocabulary differences (lift vs elevator, lorry vs truck, return ticket vs round-trip) come up naturally in scenarios; the report flags them when relevant. If you're traveling to a specific region, mention it in onboarding and the AI will calibrate.
Yes: dietary restrictions ('I'm allergic to peanuts'), asking what's in a dish, splitting the bill, asking for tap water vs bottled, complaining politely about a dish, tipping conventions. Most travelers find the dietary-restriction conversation the hardest because the vocabulary is specific and the consequences of miscommunication are real.
The AI calibrates to context you give it. Mention you're going to Tokyo and the scenarios will skew to navigating the JR system, asking about vegetarian options, etc. Mention Paris and the scenarios shift to métro tickets, café etiquette. The vocabulary cues come from your responses, the AI mirrors them.