Airbnb Coding Interview Questions

Airbnb coding interview questions are standard patterns framed in product language, and the rule that defines the round is that your code has to compile and run live in CoderPad. Graph traversal and dynamic programming are over-indexed, the signature problems (two Premium questions) hide inside booking-and-listing stories, and a Code Review round often replaces the second coding interview.

One rule sits above everything else in an Airbnb coding round: the code has to run. You write a real solution in CoderPad, hit run, and the interviewer watches it pass or fail against the test cases. No pseudocode, not at any stage. That one rule reshapes how you prep this whole section. The problems lean practical, applied over algorithmic, and they reward clean code you can debug on the spot over a flash of insight. Most are standard patterns in Airbnb-flavored framing, with a handful of craft problems where the edge cases and index discipline decide who passes.

Airbnb leans practical on purpose. The prompts give it away. Combine two listings whose date ranges cover a requested stay. Price a menu to hit an exact budget. Paginate search results without repeating a host on one page. These look like booking, search, and listings work because Airbnb deliberately picks problems that resemble the job. The difficulty underneath is real, roughly one hard problem per round, and you are not required to fully solve it. Clean partial progress with sound reasoning passes. The bar is working, readable code shipped under time, plus the composure to fix it live when the first run breaks.

The Airbnb angle

Half your grade is what happens after you hit run. You compile, the first execution comes back red, and the interviewer is reading your face while you read the error. That is the moment. Candidates who only ever solved on paper freeze here, because they have never debugged their own output with someone watching. So practice in a real editor, in the language you will use on the day, until narrating a fix feels routine. Airbnb assumes more language fluency than most big-tech loops do, and clever-but-broken loses to plain-but-running, every time.

The Airbnb Code Review Round

The Airbnb Code Review round, added around 2024, replaced the second coding interview in many loops. You get a repository of about three pull requests and an hour to review them the way you would for a teammate, often in a language you did not pick. You write no code. You read the diff, find the real bugs, and rank what matters over the nitpicks. No other big-tech loop runs a round quite like it.

What candidates do not get told up front is that you score points for the issues you raise, and the biggest pull request is worth the most. The early PRs hold the easy points, like a wrong enum value that ships a bug, or a dict lookup that should be dict.get to dodge a KeyError. Others are a missing Optional on a type hint, or swallowed exceptions that should be logged and re-raised. The large PR is where the level signal lives. A batching change to an overloaded API hides a non-thread-safe dequeue that can send the same request twice, and a queue that drops data after it fills. It also has sequential writes that should be a bulk upload, and magic constants scattered through the file. Spotting the concurrency and data-loss bugs is what separates a staff-level review.

The Airbnb angle

Budget your hour against the points rather than the line numbers. The instinct is to start at the top and polish PR one, which is backwards. Skim all three first, then give the largest PR most of your time, because that is where the architecture comments and the highest-value bugs sit. Triage as you read: call the correctness and concurrency problems first, the data-loss and security ones next, and leave style nitpicks for last or for a quick batch at the end. The round is also reading how you give feedback. Frame comments as a question or a suggestion a teammate would want rather than a verdict, because Airbnb is grading collaboration here as much as your eye for a bug.

What the coding rounds favor

Airbnb keeps the components of a normal loop but weights them unusually. Graph traversal is the largest single family, and dynamic programming is heavier than at most big-tech shops, heavy enough that recent candidates report being caught off guard by it. Around that core sits the applied band: a craft tier of simulation and parsing problems that are medium on paper and punishing on edge cases, plus the signature problems framed in product language. Strings and sliding window carry the phone screen, and intervals show up through scheduling and a Premium question family. The exact number of pure-coding rounds is in flux since the Code Review round arrived, so confirm your loop's shape with your recruiter rather than assuming the older two-algorithm format.

Read the chart as where your first hours go rather than a ranking to memorize. Graphs lead, the array-and-string fundamentals back the screen, and dynamic programming earns more weight than its share suggests because it is the family candidates most often underprepare. The six lessons in this section run in roughly that order. Each opens with how Airbnb frames the family and the variants it favors, then works the highest-return problems end to end as runnable code with the follow-up the interviewer will layer on. Where a lesson rests on a base data structure, the full walkthrough lives in the free Foundations course so these pages can stay on what is specifically Airbnb.

The Airbnb angle

Weight your prep toward DP and the applied problems, because that is where strong coders get surprised. A generic LeetCode grind covers the graph and string core fine, but it underweights the two families Airbnb over-indexes. Dynamic programming shows up more than almost anywhere else, and the signature problems like [a Premium question](/guides/companies/airbnb/coding/signature-archetypes) and a Premium question do not announce themselves as DP behind the booking-and-listing story. The other trap is treating the simulation and parsing problems as easy. Two Premium questions look medium and then bury you in off-by-one and boundary cases, which is exactly the index discipline Airbnb is screening for. Spend real hours there, with the code running.

How these lessons are organized

The six lessons move from the largest family to the Airbnb-specific capstone. Start with graphs and grids, the biggest cluster and the one the screen most often draws from, then work down through the patterns toward the signature archetypes that no pure-pattern list captures.

Common questions