How to Use ChatGPT to Find Cheap Flights

Cheap airfare isn’t luck—it’s system + flexibility. You’ll combine (1) Google Flights tools for price visibility with (2) targeted ChatGPT prompts that compress hours of search into a few smart queries and (3) traveler protections (24-hour cancels, passenger-rights rules) to lock savings without risk.

The 7-step workflow

1) Set the “deal shape” with ChatGPT

Prompt:

“I’m flying [city] → [city] with ±3 days flexibility sometime in [month]. My hard constraints: [nonstop? max 1 stop? depart after 10 AM?]. Suggest the 3 cheapest windows (weekday patterns + shoulder periods), 2 backup airports on each end, and the likely floor price based on recent trends.”

Why this works: you get an initial target fare and a map of flexible windows before you even touch a calendar.

2) Build a flexible calendar plan

Prompt:

“Create a flex calendar plan for [origin(s)] ↔ [destination(s)] over [date span]: list five outbound/inbound date pairs that usually price low (Tue/Wed/Sat departures, midweek returns), and mark blackout weeks (school breaks, big events).”

Now you know which days to try first when prices swing.

3) Use Google Flights like a pro (then feed results back)

Use Date Grid / Price Graph to surface cheapest combos fast.
Turn on Price Tracking for your route and flexible dates.
For “I’m open to anywhere,” use Explore Map with your date range and a price ceiling.
Try Flight Deals (the AI “describe your trip” tool) to surface bargains when you’re destination-flexible.

Then paste findings to ChatGPT:

“Here are the three cheapest date pairs and carriers I’m seeing. Which is the best value after bag/seat costs and connection risk? Suggest the top 2 to book and the walk-away price.”

4) Alternate airports, open-jaw, and positioning plays

Prompt:

“From [YOW/YYZ/YUL] to [target region] in [month], list realistic alternates (e.g., YTZ or YHM for Toronto area; BOS/JFK/EWR for U.S. gateways; LGW/STN for London). Propose two open-jaw plans (e.g., fly into City A, home from City B) that often price below round-trip. Flag when a positioning flight is worth it, and warn me if the savings don’t survive bags/transfer risks.”

Use this to unlock routes locals miss—especially from Ottawa (YOW) where hopping to YUL or YYZ can widen options.

5) Price-drop safety net (24-hour rule + alerts)

Many itineraries touching the U.S. are covered by a 24-hour free-cancel window when booked direct with the airline and purchased ≥7 days before departure. That lets you lock a good fare, keep tracking for a day, and rebook if it drops.
Ask ChatGPT to time-box you:

“I just booked [flight] at $X. For the next 20 hours, remind me when tracked fares go below $X-$25 so I can cancel and rebook.”

6) Total-trip math: avoid the ‘fake cheap’ ticket

Give ChatGPT the fare and fees:

“Compare these two options after bags, seat selection, connection risk, and airport transfer costs. Assume 1 carry-on + 1 checked. Price out total trip cost and rank by hassle-adjusted value.”

This stops you from “saving” $40 while paying $80 in bag/seat fees or losing a day to bad connections.

7) Book clean, keep leverage

Prefer one ticket for all legs (single PNR) to retain misconnection protection.
If you must self-transfer (separate tickets), schedule a long layover, same-terminal if possible; carry on only.
Screenshot fare rules and hold/cancel deadlines. Store your fare baseline so ChatGPT can keep a simple “rebook if under $X” rule.

Prompts that consistently find savings

A) Cheapest month vs exact week

“For [city pair] sometime [season], is it cheaper to go earlier/later? Give three 4-day windows with the best odds of sub-$[target] fares and explain why (demand patterns, shoulder season).”

B) Nonstop tax

“Price gap sanity check: How much does nonstop typically add on [route]? If I allow 1 stop, what hubs improve odds of a cheaper fare without big misconnection risk?”

C) Bag math

“List expected bag fees for [airline/cabin] on [route] and the cheapest way to bring 1 carry-on + 1 checked (fare family vs credit-card benefit vs paying at airport). Build the all-in price.”

D) Separate tickets risk rating

“I’m considering [itinerary with self-transfer]. Give a risk score (missed connection, immigration, terminal change, weather) and a minimum safe layover.”

E) Nearby airport tradeoff

“From [home], quantify cost/time to use [alternate airports] including ground transfer. What’s my break-even flight savings to switch airports?”

Tactics people forget (that quietly save $100+)

Shift your return by one weekday. Even a Sat-night return often undercuts Sunday.
Mix-and-match carriers on round-trips within one ticket (many search engines can pair different airlines).
Open-jaw into multi-city trips: fly into a cheap hub, train/bus to your real destination, fly home from there.
Avoid basic economy if you need a carry-on/seat—“cheapest” isn’t cheap after add-ons. (Google Flights now lets you exclude basic economy in filters.)
Southwest fares show in Google Flights now, but you still book on the airline’s site—include them when pricing U.S. routes.
Mistake-fare mindset: set broad destination alerts and keep a small emergency fund; when a unicorn appears you can act fast, then cancel within your 24-hour window if needed.
Use schedule changes to your advantage: if the airline materially shifts your flight time later, you can often move to a better flight for free—ask ChatGPT to draft the request.

Rights & rules that protect your wallet (know these)

U.S. 24-hour rule: On most tickets to/from/within the U.S., airlines must hold or refund within 24 hours (requirements vary by whether payment is taken; book direct and ≥7 days out to be safe). Use this to lock fares while you compare.
Canada (APPR): Delays/cancellations within the airline’s control can trigger compensation (amount depends on airline size and delay length). File in writing; airlines have 30 days to respond.
EU/EEA/UK (EU261/UK261): Strong cash compensation for long delays/cancellations when the operating carrier is EU/UK or the flight departs the EU/UK; weather/ATC are exceptions.
Skiplagging (hidden-city) warning: Airlines hate it; you can lose miles or get cancelled if caught. If you ever do it, carry-on only and one-ways—but know it violates fare rules and may invite hassle.

Canada-based examples (Ottawa focus, but apply broadly)

Alternate airports that often save:

From YOW, compare YUL (Montréal) and YYZ (Toronto Pearson). For some U.S. trips, YUL can be cheaper midweek; for sun destinations, YYZ has more competition.
If you’re Toronto-area, include YTZ (Billy Bishop) and YHM (Hamilton) for select routes.

Seasonality you can exploit:

Shoulder seasons (late Apr–May; mid-Sep–early Nov) are prime for Europe and Caribbean deals.
Avoid March Break and major holiday weeks unless you book months early.

ChatGPT prompt to localize:

“From Ottawa to Lisbon in late September, list the 3 cheapest date pairs historically, suggest whether to route via YUL/YYZ/BOS, and give a walk-away price after typical bag costs.”

Example: putting it all together in 9 minutes

Ask for deal shape (Step 1).
Build a flex calendar plan (Step 2).
On Google Flights, check Date Grid + Price Graph, enable tracking, peek at Explore and Flight Deals (Step 3).
Paste results; have ChatGPT do total-trip math with bags/seat fees (Step 6).
Book direct, screenshot fare rules, and set a 24-hour “rebook if cheaper” reminder (Step 5 & 7).

Troubleshooting (why your “cheap” fare isn’t cheap)

Basic economy traps: No full-size carry-on or no seat choice; paying à la carte often beats booking the rock-bottom fare.
Tight self-transfers: Two tickets = no protection. Miss a leg and you may buy a walk-up fare.
Wrong airport math: A $60 cheaper fare out of another city becomes $120 more after transit/parking/time.
Bag math blindness: One airline’s slightly higher fare with free carry-on + personal item often wins.

Ask:

“Audit this itinerary for hidden costs (bags, seat, airport transfer, self-transfer risk). Give me the true total and a go/no-go.”

One-week plan (from search to booked)

Day 1: Run deal-shape prompt; set tracking on 2–3 date pairs.
Day 2: Explore alternates (airports, open-jaw).
Day 3: Check alerts; if within $25 of target, price total trip with bags and book.
Day 4: Use 24-hour window to re-check and rebook if lower.
Day 5: If not booked, widen dates by ±2 days and include one new alternate airport.
Day 6: Re-assess “walk-away price.” If trend rises, pull trigger.
Day 7: Lock seats/bags strategically (often cheaper online than at airport).

Quick prompt kit (paste into ChatGPT)

Deal shape: “Cheapest 3 windows + target fare for [origin]→[dest] in [month], with [constraints].”
Alternates: “Best alternates for [region] including open-jaw/positioning ideas. Add risk notes.”
Total cost: “Compare Option A vs B after bags/seats/ground transfer. Rank by value.”
24-hour guardrail: “I booked at $X. For 20 hours, alert me if tracked fares fall below $X-$Y.”
Self-transfer risk: “Rate this connection risk and minimum safe layover: [airport/itinerary].”

TL;DR (finally)

Use ChatGPT to scope when/where flights are cheapest, then verify on Google Flights with Date Grid/Price Graph/Explore and tracking.
Consider alternate airports, open-jaw, and mix-carrier options—but do the all-in price with bags and transfer time.
Book direct and leverage the 24-hour free-cancel rule (when eligible) to hold a fare while you keep tracking.
Know your passenger rights (APPR, EU261/UK261) and the risks of skiplagging.
The goal isn’t the cheapest sticker price—it’s the lowest total cost with the least hassle.