Interview experience
Meta Interview Questions and Answers (2026)
Overview
Section titled “Overview”Meta’s loop runs a recruiter screen, an optional CodeSignal assessment, a phone screen, and a 4-round onsite - coding, a new AI-enabled coding round, system design, and behavioral - hired mainly through referrals and a small early-career track, not a campus drive.
Meta interview process at a glance
Section titled “Meta interview process at a glance”| Round | Duration | What it tests |
|---|---|---|
| Recruiter screen | 20-30 min | Background, motivation, logistics |
| Online Assessment (CodeSignal, some candidates) | 70-90 min | Progressive multi-stage coding problem, video/mic monitored |
| Technical phone screen | 45 min | 1 live coding problem |
| Onsite - Coding (traditional) | 45 min | DSA - arrays/strings, graphs, trees, intervals |
| Onsite - Coding (AI-enabled, since late 2025) | ~60 min | Coding with a built-in AI assistant; your judgment and prompting, not the assistant’s output |
| Onsite - System design / architecture | 45-60 min | Scaled to level - lighter architecture chat for E3, full distributed design for E4+ |
| Onsite - Behavioral | 45 min | 5 signal areas: driving results, ambiguity, communication, growth, conflict resolution |
| Team matching | Varies | Team/org placement after the offer |
Recruiter screen
Section titled “Recruiter screen”A 20-30 minute call to confirm background, work authorization, timeline, and why Meta. It’s a logistics filter, not a technical bar - be direct about your experience level and availability.
Common questions
- Walk me through your resume/current role
- Why Meta, and why now?
- What team or org are you interested in?
- What’s your notice period / earliest start date?
Full behavioural frameworks are on the Meta HR interview questions page.
Online Assessment (CodeSignal)
Section titled “Online Assessment (CodeSignal)”Added to Meta’s process during 2025, this is a single problem split into 4 progressively harder stages (for example, building an in-memory key-value store, then adding TTL, then point-in-time reads, then deletion) inside a 70-90 minute video/mic-monitored CodeSignal session. Not everyone sees this stage - plenty of candidates, especially referrals, get routed straight to the phone screen.
Common questions
- Stage 1: implement basic get/set operations on an in-memory store
- Stage 2: add TTL (time-to-live) expiry to entries
- Stage 3: support point-in-time or historical reads
- Stage 4: add deletion and other advanced operations
See how real candidates handled this stage on the Meta interview experience page.
Technical phone screen
Section titled “Technical phone screen”One 45-minute call with an engineer: a single live coding problem, occasionally two shorter ones. Interviewers weigh clarifying questions and complexity discussion alongside a working solution.
Common questions
- Two Sum / Valid Parentheses (warm-up pairs for lighter loops)
- Longest Substring Without Repeating Characters
- Minimum Remove to Make Valid Parentheses
- Explain the time/space complexity of your approach
Onsite: coding rounds
Section titled “Onsite: coding rounds”Historically two traditional coding rounds; since roughly Q4 2025 many SWE loops swap one for the new AI-enabled coding round - a ~60-minute session in a coding environment with a built-in AI assistant. You’re still evaluated on your own engineering judgment, so candidates report writing a first pass unassisted, then using the tool to catch missing edge cases and narrating why a prompt did or didn’t help.
Common questions
- Merge Intervals
- Binary Tree Vertical Order Traversal
- Number of Islands
- Group Anagrams
- Product of Array Except Self
- Diameter of a Binary Tree
Round-by-round breakdowns with the exact follow-ups asked are on the Meta interview experience page.
Onsite: system design / architecture
Section titled “Onsite: system design / architecture”Depth scales sharply with level: E3 loops often get a lighter architecture discussion instead of a full design round, while E4/E5 candidates get a genuine distributed-systems design built around Meta’s own product surface.
Common questions
- Design Facebook/Instagram News Feed (ranking, caching, real-time updates)
- Design Messenger or a chat/notification system at scale
- Design a search or content-ranking service
- Discuss consistency vs. availability trade-offs for your design
Onsite: behavioral round
Section titled “Onsite: behavioral round”Community reports describe this round as feeling informal - almost a “vibes check” on communication and presentation - but it carries real weight: Meta scores answers against five signal areas (driving results, embracing ambiguity, effective communication, continuous growth, and conflict resolution), and unlike coding questions, interviewers aren’t working from a fixed script.
Common questions
- Describe the highest-impact project you’ve shipped - how did you measure that impact?
- Tell me about a time you had to decide with incomplete information, or move fast on something that later broke
- Tell me about a time you gave direct, difficult feedback to a teammate or manager
- Tell me about a time you failed and what you changed afterward
- Tell me about a disagreement with your manager or a teammate and how you resolved it
Sample answer frameworks for each of these are on the Meta HR interview questions page.
Leveling: why design and behavioral rounds decide E4 vs E5
Section titled “Leveling: why design and behavioral rounds decide E4 vs E5”This is the part of Meta’s process candidates most often get wrong. Coding rounds function mostly as a pass/fail gate, but public discussion (Blind threads, senior-engineer interview guides) consistently says the system-design and behavioral interviewers carry the most weight in the final hire-and-level call - to the point that strong coding performance doesn’t protect you from a down-level, or even a “no hire,” if the design or behavioral signal comes back weak for your target level. There’s no published Google-style Hiring Committee or Amazon-style Bar Raiser at Meta; the panel of interviewers discusses feedback together, and a split decision can trigger one more follow-up interview for extra signal. Treat this as a strong, recurring pattern from candidate reports rather than Meta’s official policy, since Meta does not publish its internal leveling rubric.
Common technical interview questions and answers
Section titled “Common technical interview questions and answers”Q: How do you solve Merge Intervals?
Sort the intervals by start time, then sweep once keeping a current interval. For each next interval, if its start is at most the current interval’s end they overlap, so extend the current end to the maximum of the two ends; otherwise push the current interval to the result and make the next one current. Push the final interval after the loop. Sorting dominates at O(n log n) time with O(n) output space. The follow-ups Meta interviewers reach for are inserting a single new interval into an already-sorted list, which is O(n) with a three-phase scan, and handling touching intervals such as [1,3] and [3,5], where you must ask whether they should merge.
Q: Minimum Remove to Make Valid Parentheses - what is the approach?
Do two passes with a stack of indices. In the first pass, scan left to right: push the index of each open bracket; on a close bracket, pop if the stack is non-empty, otherwise record that close bracket’s index as unmatched. After the pass, whatever indices remain on the stack are unmatched open brackets. Collect both sets, then build the result string skipping those indices. That is O(n) time and O(n) space, and it removes the minimum number of characters because every index marked was provably unmatchable. A common mistake is deleting from the string during the scan, which shifts indices - mark first, rebuild once.
Q: How do you do a vertical order traversal of a binary tree?
Run a BFS while carrying a column index: the root is column 0, a left child is column minus one, a right child is column plus one. Append each node’s value into a hash map keyed by column, and track the minimum and maximum column seen. Because BFS visits level by level, values land in top-to-bottom order automatically, which is why BFS is preferred over DFS here - a DFS needs an explicit sort by row afterwards. At the end, emit the columns from minimum to maximum. Time is O(n) and space O(n). The clarifying question worth asking is how to order two nodes in the same row and column: the harder variant requires sorting those values, the classic Meta version accepts left-to-right insertion order.
Q: How would you count the number of islands in a grid?
Scan every cell; when you find an unvisited land cell, increment the island counter and flood-fill from it with DFS or BFS, marking every reachable land cell visited so it is never counted again. Each cell is visited a constant number of times, so it is O(rows times columns) time; space is O(rows times columns) in the worst case for the recursion stack or queue. Mutating the input grid in place (writing a 0 over visited land) avoids a separate visited matrix, but you should say out loud that you are modifying the caller’s data and offer to restore it. Typical follow-ups are switching to 8-directional connectivity, handling a grid too large for memory by streaming rows, and the union-find variant used when islands are added one at a time.
Q: Product of Array Except Self, without division.
Build the answer from two directional prefix products. First pass left to right: result[i] holds the product of everything before i, maintained by a running variable starting at 1. Second pass right to left: keep a running suffix product, multiply it into result[i], then update it with the original nums[i]. That gives O(n) time and O(1) extra space beyond the output array, which is what the interviewer is looking for. Division is banned partly because a single zero anywhere breaks it; the two-pass approach handles zeros naturally, since one zero leaves exactly one non-zero entry and two zeros make everything zero.
Q: Longest Substring Without Repeating Characters - what is the optimal solution?
Use a sliding window with a hash map from character to its last seen index. Move the right pointer across the string; when the current character was seen at an index at or after the current left pointer, jump the left pointer to one past that index rather than stepping it one at a time. Update the best length as right minus left plus one on every iteration. Each character is processed once, so it is O(n) time and O(min(n, alphabet size)) space. The jump is what makes it a single pass - the naive version that shrinks the window one character at a time is still correct but reasons its way towards O(n squared) in the discussion.
Q: How do you compute the diameter of a binary tree?
The diameter is the longest path between any two nodes, measured in edges, and it need not pass through the root. Write a recursive height function that returns the height of a subtree, and while it unwinds, update a shared maximum with leftHeight plus rightHeight, which is the length of the longest path bending at that node. Return one plus the maximum of the two child heights to the parent. That single post-order traversal is O(n) time and O(h) stack space, where h is the tree height - O(log n) balanced, O(n) in a skewed tree. Computing height separately at every node instead would be O(n squared), and getting the edges-versus-nodes definition wrong by one is the most common slip.
Q: How would you design Facebook News Feed?
Split it into write path and read path. On write, a post goes to a post store and a fanout service pushes the post id into each follower’s precomputed feed list in a fast store such as Redis - fanout-on-write - so a read is a single cheap range query. That breaks for celebrity accounts with tens of millions of followers, so those are handled fanout-on-read: their posts are fetched at request time and merged into the feed, giving the hybrid model that is the expected answer. Ranking then scores candidate posts on recency, affinity between viewer and author, and predicted engagement, rather than showing pure reverse-chronological order. Round out the design with cursor-based pagination rather than offsets, aggressive caching of the top of each feed, and eventual consistency - a post appearing a few seconds late is acceptable, whereas a lost post is not, so fanout runs through a durable queue.
Frequently asked questions about Meta interviews
Section titled “Frequently asked questions about Meta interviews”What is the Meta interview process?
Meta’s loop usually runs: 1. Recruiter screen (20-30 min) - background, motivation, logistics. 2. Online Assessment (CodeSignal, ~70-90 min, video/mic monitored) - not every candidate gets this; it’s a progressive multi-stage coding problem added to the process in 2025. 3. Technical phone screen (45 min) - one live coding problem. 4. Virtual onsite (4 rounds) - one traditional coding round, one AI-enabled coding round (rolled out since late 2025, with access to a built-in AI assistant), one system design/architecture round, and one behavioral round. 5. Team matching after the offer. Total timeline is roughly 3-4 weeks, though it varies by team and level.
What questions are asked in Meta interviews?
Coding rounds lean on arrays/strings, sliding window, graphs (BFS/DFS), trees, and interval problems - Merge Intervals, Binary Tree Vertical Order Traversal, and Minimum Remove to Make Valid Parentheses are frequently reported. System design covers Meta’s own products (News Feed, Messenger, notification systems) at a depth scaled to level. The behavioral round is scored against five signal areas: driving results, embracing ambiguity, effective communication, continuous growth, and conflict resolution.
What is Meta’s AI-enabled coding round?
Since roughly Q4 2025, Meta has been replacing one of the two onsite coding rounds with an AI-enabled coding round on many SWE and EM loops, reportedly up through senior levels. You get a coding environment with a built-in AI assistant for about 60 minutes; interviewers are still evaluating your own engineering judgment rather than the assistant’s output, so candidates report writing a first pass themselves, then using the assistant to review edge cases and explaining why a given prompt did or didn’t help. Treat this as a live rollout - not every loop has it yet, so confirm with your recruiter.
How does Meta decide between E4 and E5, or down-level a candidate?
Community reports (Blind, interview-prep guides) consistently say Meta’s coding rounds mostly gate pass/fail, while the system design and behavioral rounds carry the most weight for your final level. A candidate can clear coding cleanly and still be down-leveled - or rejected - on a soft design or behavioral round, especially at E4/E5. Treat this as a directional pattern from public discussion, not an official Meta policy, since Meta doesn’t publish its internal leveling rubric.
How many rounds are there in the Meta interview?
Typically 5-6 touchpoints: a recruiter screen, an optional CodeSignal online assessment, a technical phone screen, and a 4-round onsite (coding, AI-enabled coding, system design, behavioral), followed by team matching after the offer. Not every candidate sees the online-assessment stage - many go straight from recruiter screen to phone screen.
How should I prepare for Meta interviews?
Drill medium-hard DSA (arrays/strings, graphs, trees, intervals) until you can explain trade-offs while coding, practice a system-design answer scaled to your target level, and prepare distinct STAR stories mapped to Meta’s five behavioral signals - driving results, embracing ambiguity, communication, growth, and conflict resolution - since interviewers can ask about any of them without a fixed script.
How selective is Meta’s India hiring compared to a campus drive?
Far more selective. Meta’s India entry point is a University Grad / early-career program (mainly Bengaluru) plus a summer internship track - both much smaller than a TCS- or Infosys-style drive - with most hiring flowing through targeted postings, referrals, or off-campus applications rather than a college-wide placement cell. Expect the same round structure as experienced hires, just with the design round scaled lighter for E3.