Meta Coding Interview Questions: Patterns and Top Problems
Meta interviews reward speed. The question list below proves it. Most tagged problems are classics you have seen before, and the rounds are short. Candidates who recognize the pattern in the first minute do well. Candidates who search for it mid-round do not.
This page gives you the target list. It ranks the coding questions that appear most often in Meta’s tagged problem data, and each row links to the pattern guide on this site that teaches the technique. It ends with a study order matched to how Meta actually runs its loop.
The short version
- Medium holds 51 percent of tagged problems. Easy holds 38 percent. Hard holds 11 percent.
- Arrays touch about 48 percent of tagged problems. Strings follow at about 22 percent.
- Two pointers cover about 18 percent, nearly double the Google share. The top 20 rewards standard templates known cold.
Meta’s Interview Loop
- A recruiter call covers your background and level.
- One technical phone screen follows. You write runnable code in CoderPad.
- Phone screens often contain two problems in 40 to 45 minutes.
- The onsite loop runs several coding rounds across one or two days.
- Each coding round often contains two problems.
- Senior levels add system design or product sense rounds.
- A behavioral round checks how you work with other people.
Two problems per round is the detail to prepare for. Interviewers expect working code fast, and many candidate reports say a slow start costs the round even when the final solution is correct. Practice under a timer from the start. Recognition speed is a trainable skill, not a personality trait.
The Questions Meta Asks Most
LeetCode tags problems by company when candidates report them. Every tagged problem carries a relative frequency score from 0 to 100. The 20 below score highest for Meta right now, and 9 of them sit in the very top band. Each row links to our pattern guide for that problem’s technique.
If you learn five guides first, learn these
- Start with two pointers . Four rows of the top 20 run on it.
- Study sliding window alongside it. It covers Longest Substring Without Repeating Characters and Minimum Size Subarray Sum.
- Take array manipulation third. Arrays touch about 48 percent of tagged problems.
- Bring in hash table for Two Sum and Group Anagrams.
- Finish with Kadane’s algorithm for Best Time to Buy and Sell Stock and Maximum Subarray.
| Problem | Difficulty | Demand | Pattern Guide |
|---|---|---|---|
| Two Sum | Easy | Very high | Hash Table |
| Longest Substring Without Repeating Characters | Medium | Very high | Sliding Window |
| Subarray Sum Equals K | Medium | Very high | Prefix Sum |
| Add Two Numbers | Medium | Very high | Linked List |
| Palindrome Number | Easy | Very high | Math Algorithms |
| Median of Two Sorted Arrays | Hard | Very high | Binary Search |
| Best Time to Buy and Sell Stock | Easy | Very high | Kadane’s Algorithm |
| Kth Largest Element in an Array | Medium | Very high | Heap Priority Queue |
| Group Anagrams | Medium | Very high | Hash Table |
| Remove Duplicates from Sorted Array | Easy | High | Two Pointers |
| Move Zeroes | Easy | High | Array Manipulation |
| 3Sum | Medium | High | Two Pointers |
| Valid Parentheses | Easy | High | Stack |
| Next Permutation | Medium | High | Array Manipulation |
| Minimum Size Subarray Sum | Medium | High | Sliding Window |
| Maximum Subarray | Medium | High | Kadane’s Algorithm |
| Reverse Integer | Medium | High | Math Algorithms |
| Longest Palindromic Substring | Medium | High | Two Pointers |
| Search in Rotated Sorted Array | Medium | High | Binary Search |
| Container With Most Water | Medium | High | Two Pointers |
Read this table against the Google one and a difference jumps out. Meta’s list is shorter and more concentrated at the top. Nine of these 20 sit in the highest frequency band, and almost every entry is a standard problem with a known template. At Meta, knowing the standard solution cold matters more than improvising something new.
Topics and Difficulty at Meta
The most common topics among all tagged Meta problems are below. The share is the percent of tagged problems that carry the topic. One problem can carry several topics, so the shares add to more than 100 percent.
| Topic | Tagged Problems | Share |
|---|---|---|
| Array Manipulation | 95 | ~48% |
| String Manipulation | 43 | ~22% |
| Hash Table | 38 | ~19% |
| Two Pointers | 35 | ~18% |
| Math Algorithms | 30 | ~15% |
| Dynamic Programming | 28 | ~14% |
| Sorting | 24 | ~12% |
| Binary Search | 20 | ~10% |
| Database (SQL) | 16 | ~8% |
| Linked List | 15 | ~8% |
Sorting has no link because we treat it as part of other patterns rather than a pattern of its own. Database has no link because SQL is outside our pattern guides. Both matter less for prep than the rows above them. Arrays, strings, hashing, and two pointers together cover most of what Meta asks, and two pointers carries nearly double the share it does at Google. Dynamic programming is lighter here than at most big companies. Learn the standard subarray and string cases, then spend your remaining time on speed instead of harder DP variants.
Difficulty splits cleaner than you might expect:
| Difficulty | Problems | Share |
|---|---|---|
Easy | 75 | 38% |
Medium | 101 | 51% |
Hard | 21 | 11% |
Meta asks more Easy questions than Google does, and fewer Hards. That fits the two-problems-per-round format, because one Hard per round would leave no time for a second problem. Do not skip Easy practice at Meta. A slow Two Sum costs as much signal as a failed Medium.
Method: LeetCode company-tagged lists for Meta, ranked by relative frequency scores from 0 to 100. Shares overlap because one problem can carry several topics.
Where to Spend Your Prep Time
- Master arrays, strings, and hash tables before anything else. They cover about half of all tagged problems. The array manipulation , string manipulation , and hash table guides hold the base techniques.
- Pair two pointers with sliding window. Two pointers alone covers nearly one tagged problem in five. The two techniques share mechanics, so learn them side by side in the two pointers and sliding window guides.
- Own the subarray family completely. Best Time to Buy and Sell Stock, Maximum Subarray, and Subarray Sum Equals K all rank in the top 20. Study the Kadane’s algorithm and prefix sum guides until you can recite both templates.
- Practice two problems in 40 minutes, every week. This matches the real round format better than single-problem sessions. Use two Mediums, or one Easy plus one Medium.
- Keep everything reviewed on a schedule. Speed comes from recall, and recall fades. Spaced repetition brings each problem back just before you would forget it.
graph TD
A["Array, string, and hash basics"] --> B["Two pointers plus sliding window"]
B --> C["Subarray templates"]
C --> D["Two problems in 40 minutes"]
D --> E["Top twenty solved under time"]
E --> F["Scheduled recall until the phone screen"]
The diagram walks the prep plan from base patterns to timed rounds and review mode.
Questions Candidates Ask
How bad is the time pressure really?
Most coding rounds hold two problems. Phone screens often fit two problems into 40 to 45 minutes. Practice under a timer from the start, since recognition speed is trainable.
Why does Meta ask so many Easy questions?
Easy problems hold 38 percent of the tagged set. Rounds with two problems leave no room for a Hard plus a second question. Treat Easy speed as scored signal, not warmup.
Which guide should I open first?
Open the array manipulation guide. Arrays touch about 48 percent of tagged problems. Then study two pointers with sliding window .
How much dynamic programming does Meta need?
Less than most big companies. Dynamic programming sits at about 14 percent of tagged problems. Learn the standard subarray and string cases in the Kadane’s algorithm guide, then spend the rest on speed.
Next Steps
Work through the top 20 list twice. First pass for understanding, second pass under a timer with the linked pattern guides closed. The coding patterns index lists every guide, and sliding window plus two pointers are the best first reads for Meta specifically. Comparing companies? See how the Google , Amazon , and Apple guides differ on topic weight and difficulty.