Meta Interview Guide

Meta’s coding rounds usually pack two problems into 45 minutes, which changes the optimisation target completely. At Google you go deep on one problem; here you must finish two, and running out of time on the second is the most common way strong candidates fail.

That makes fluency the binding constraint rather than insight. The problems are drawn from a relatively stable pool of pattern families — the list below is the durable signal — and the expectation is that you recognise, state, and implement without long pauses.

Crowd-reported, not published

No company publishes its question bank. The round structure and the bar below are corroborated by recruiters and candidates; the problem list is crowd-reported, stale, and biased toward whoever posted it. Treat the pattern families as the durable signal — which families a company leans on changes far more slowly than which individual problems it uses.

Speed and volume. Two problems in 35-40 minutes of coding time is the signature format, so a pattern you have to derive from scratch will cost you the round even if you get there.

The loop

Phone screen 1× · 45 min Two mediums back to back

Coding (onsite) 2× · 45 min Two problems each, mediums with one hard

Behavioural ("Jedi") 1× · 45 min Conflict, ownership, moving fast

System design 1× · 45 min E5 and above

The bar

Interviewers are looking for a working solution plus a stated complexity inside ~20 minutes per problem. Optimal-first matters more here than anywhere else — there is no time to iterate from brute force.

What they lean on

This is the durable part. Which pattern families a company favours is far more stable than which individual problems it uses, so prepare in this order:

1. two-pointers
2. tree-dfs-paths-and-sums
3. binary-search-template-and-variants
4. hash-tables
5. grid-traversal-islands-and-flood-fill
6. one-dimensional-dp
7. stack-parsing-and-expression-evaluation

Quirks worth knowing

• Budget 20 minutes per problem and say so out loud. Ask which problem to start with if given a choice.
• Meta reuses a comparatively small rotation of problems; the "top 50 Meta-tagged" lists have unusually high hit rates.
• Clarify input assumptions in one sentence, not five. Over-clarifying eats the clock you cannot spare.
• You are hired to the company, then matched to a team afterwards, so breadth beats depth in any one domain.

Reported problems

Read this before you trust the list. No company publishes its question bank. Every "asked at Meta" list in circulation — including this one — is aggregated from candidate reports, is months to years out of date, and over-represents whoever bothered to post. Use it to choose an order to practise in. Do not use it as a question bank, and do not assume a problem missing from it is safe to skip.

29 problems

5 easy19 medium5 hard

AllEasyMediumHard

Unsolved onlyExportImportReset

• ○1Two SumeasyNeetCode 150Blind 75LeetCode Top Interview 150
• ○20Valid ParentheseseasyNeetCode 150Blind 75LeetCode Top Interview 150
• ○121Best Time to Buy and Sell StockeasyNeetCode 150Blind 75LeetCode Top Interview 150
• ○206Reverse Linked ListeasyNeetCode 150Blind 75
• ○242Valid AnagrameasyNeetCode 150Blind 75LeetCode Top Interview 150
• ○3Longest Substring Without Repeating CharactersmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○153SummediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○56Merge IntervalsmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○200Number of IslandsmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○11Container With Most WatermediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○33Search in Rotated Sorted ArraymediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○49Group AnagramsmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○75Sort Colorsmedium
• ○102Binary Tree Level Order TraversalmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○139Word BreakmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○207Course SchedulemediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○236Lowest Common Ancestor of a Binary TreemediumLeetCode Top Interview 150
• ○238Product of Array Except SelfmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○347Top K Frequent ElementsmediumNeetCode 150Blind 75
• ○133Clone GraphmediumNeetCode 150Blind 75LeetCode Top Interview 150
• ○143Reorder ListmediumNeetCode 150Blind 75
• ○380Insert Delete GetRandom O(1)mediumLeetCode Top Interview 150
• ○621Task SchedulermediumNeetCode 150
• ○739Daily TemperaturesmediumNeetCode 150
• ○42Trapping Rain WaterhardNeetCode 150LeetCode Top Interview 150
• ○76Minimum Window SubstringhardNeetCode 150Blind 75LeetCode Top Interview 150
• ○295Find Median from Data StreamhardNeetCode 150Blind 75LeetCode Top Interview 150
• ○297Serialize and Deserialize Binary TreehardNeetCode 150Blind 75
• ○269Alien DictionarypremiumhardNeetCode 150Blind 75

Pitfalls

• Spending 30 minutes on the first problem. Two problems, 45 minutes: that is roughly 20 minutes each including clarification and testing. If you are 15 minutes in with no working code, say so and simplify.
• Over-clarifying. Ask the one or two questions that change your approach, then move. A five-minute clarification phase costs a quarter of your budget for that problem.
• Hand-rolling what the stdlib provides. Writing your own heap or binary search burns minutes you do not have. heapq, bisect, Counter, deque — know them cold.
• Leaving code untested because time ran out. Trace one small input out loud. An untested solution reads as unfinished even when it is correct.

Interview follow-ups

They ask	What they’re checking	The answer
“Can you do that faster?”	Fluency, not depth	Answer in one sentence and start typing. Long silences cost more here than a slightly suboptimal first pass would.
“Let’s move on to the second problem.”	Whether you can let go	Take it. Finishing one problem and half of another is a weaker outcome than solving both adequately.
“Walk me through your code.”	Whether you can trace under time pressure	Trace it on the smallest input that exercises the interesting branch. Do this unprompted — it is fast and it is the strongest signal available.
“How would you test this?”	Engineering instinct in 30 seconds	Categories, not examples: empty, single element, all equal, and the boundary the algorithm turns on.

Recall card

• Two problems, 45 minutes. ~20 minutes each. Time management is the skill being tested.
• Fluency over insight. The pool of pattern families is stable; recognise, state, implement without pausing.
• Know the stdlib cold — heapq, bisect, Counter, deque. Hand-rolling costs minutes you do not have.
• Clarify in two questions, then move.
• Trace one small input unprompted, even when time is tight.