Skip to main content

How Meta Interviews in 2026: Process, Questions & What They Score

The complete process, reported questions, and what interviewers score.

Explore the playbookJump to the questions

Meta / Interview snapshot

8

stages from
application to offer

33

reported questions

4-6

weeks, typically

33

public sources

Last verifiedJul 2026
HOW TO READ THIS PLAYBOOK

Independent: SupaCV is not affiliated with or endorsed by Meta. This is one well-documented shape of Meta's interviews, not a script: processes change and vary by role, team, level, and region, so confirm specifics with your recruiter.

High confidenceSources
On this page

01 / For every Meta candidate

The process, start to finish.

Meta's hiring pipeline for engineering roles is a multi-stage funnel: an initial recruiter conversation, a technical phone screen, a same-day "full loop" of 4-5 onsite/virtual interviews, and a post-loop hiring-committee review followed by team matching before an offer is finalized.

On timeline: independent trackers (IGotAnOffer, interviewing.io) put the core interview stages (recruiter screen through hiring-committee decision) at roughly 4-6 weeks for engineering roles, but Meta's own careers-site FAQ states the typical end-to-end process (including sourcing and negotiation) runs "2 to 3 months," and team matching alone is commonly reported to take 2-6 weeks. Treat "a few weeks" as the optimistic case and 2-3 months as the safer planning assumption, especially if you're talking to multiple teams.

As of late 2025/2026, Meta added a proctored CodeSignal online assessment ahead of the phone screen for many (not all) candidates, and is rolling out an AI-assisted coding round (with an in-editor AI assistant, using models like GPT-5, Claude, or Gemini depending on rollout) that replaces one of the two traditional coding interviews in the onsite loop for E3-E6; and, per multiple 2026 prep-source reports, replaces coding entirely at E7+/Manager-1, where candidates get only one coding round and it is AI-assisted.

This is expanding through 2026, so confirm with your recruiter whether it applies to your loop. Unlike Amazon or the MBB consulting firms, Meta does not have one universally-cited, formally named interview rubric (like "Leadership Principles" or a case-interview format); instead, behavioral interviews (internally nicknamed the "Jedi" round) are evaluated against Meta's six publicly stated company values, which numerous interview-prep sources explicitly use as the de facto behavioral framework.

Process details (number of rounds, whether it's "system design" vs. "product design," AI-round rollout) vary by level (E3/E4 vs E5 vs E6+) and by org (infra vs product vs ML), so treat specifics as directionally accurate rather than universal.

Stage 1:

Resume/Application Review

Varies; days to a few weeks

Application or recruiter sourcing; resume is screened before any live conversation.

Stage 2:

Recruiter Screen

20-45 minutes

A lightweight, mostly non-technical call with a Meta recruiter covering background, experience, motivations, logistics, and level/comp expectations. May include a few softball behavioral questions.

Stage 3:

Online Assessment

~90 minutesNewer addition, role-dependent

A proctored CodeSignal assessment (webcam/mic/screen monitored, plus an ID check) consisting of one complex problem split into progressively unlocking stages (e.g., building a small in-memory service or key-value store). Reported as a newer gate added in 2025 for some candidate pools before the phone screen. Plenty of candidates still get routed straight to a phone screen with no OA at all.

Stage 4:

Technical Phone Screen

45 minutes

One (sometimes two) video interviews with a Meta engineer via a shared editor (CoderPad). Mostly data-structures-and-algorithms coding; usually 2 Easy/Medium LeetCode-style problems, roughly 15-20 minutes each, with a few minutes of intro/behavioral chat and Q&A. Dynamic-programming-heavy questions are reportedly de-emphasized or avoided at this stage, and getting a working algorithm matters more than a bug-free implementation on the first problem.

Stage 5:

Onsite / Virtual 'Full Loop'

Each round 45-60 minutes; full loop is roughly a half-day to full day

4-5 back-to-back interviews (sometimes compressed into one day, sometimes split across two): typically 2 coding rounds for E3-E6 (one of which is increasingly the new AI-assisted coding round with an in-editor AI assistant; multiple 2026 prep sources report this eventually replaces the second coding round entirely for E7+/Manager-1 candidates, who face only one coding interview and it is AI-assisted), 1-2 System Design or Product Design rounds (product-design leans toward API/data-flow/client-server and usability; system design leans toward distributed systems and scale), and 1 Behavioral/'Jedi' round. E6+ candidates may face an additional Leadership/organizational-impact-focused interview, are expected to pass both design rounds, and design feedback tends to carry extra weight at senior levels.

Stage 6:

Hiring Committee Review

Several days to ~1-2 weeks

Interview feedback and confidence scores are compiled and reviewed by a centralized hiring committee (not just the hiring manager) that makes the hire/no-hire and leveling decision. Coding feedback tends to drive a binary hire/no-hire signal; system/product design feedback tends to inform leveling; behavioral and design feedback reportedly carry outsized weight when a panel is split.

Stage 7:

Team Matching

1-6 weeks; frequently the single biggest bottleneck in the whole process

Since a 2023 process change, team matching happens before the formal offer: candidates have 'match' conversations with hiring managers across teams (sometimes just a couple, sometimes 10+), and both sides must opt in before an offer is extended.

Stage 8:

Offer

Days

Offer is extended contingent on successful team match; decision/negotiation windows are often short unless the candidate has competing offers.

A tag is the condition given in brackets in the stage name.

Reported interview time, one scale in minutes

Where the minutes go

  • Recruiter Screen20 to 45 min
  • Online Assessment~90 min
  • Technical Phone Screen45 min
  • Onsite / Virtual 'Full Loop'~45 to 60 min
Low end of what is reportedUp to the high end

Not drawn: Resume/Application Review, Hiring Committee Review, Team Matching and Offer, whose durations are not given in minutes.

02 / What Meta scores

The framework.

Meta's Six Company Values (used informally as the behavioral-interview evaluation lens; not a formally branded 'interview framework' the way Amazon's Leadership Principles are)

  1. 01

    Move Fast

  2. 02

    Focus on Long-Term Impact

  3. 03

    Build Awesome Things

  4. 04

    Live in the Future

  5. 05

    Be Direct and Respect Your Colleagues

  6. 06

    Meta, Metamates, Me

These six values are officially published on Meta's careers/culture page (metacareers.com/culture) and were introduced by Mark Zuckerberg in February 2022 as part of the Meta rebrand; the current page lists exactly these six (no seventh value or renamed item). Meta does NOT publicly brand these as an 'interview rubric' the way Amazon does with its Leadership Principles. There is no official Meta document instructing interviewers to score candidates against each named value.

However, multiple independent career-coaching sources (interview-prep blogs, Medium/DEV.to guides, prepfully/prachub-style guides) consistently report that Meta's behavioral round (nicknamed the 'Jedi' interview) is evaluated against these values in practice, and coach candidates to explicitly map STAR stories to them. For context on the nickname itself: a long-running, informal naming convention traced to Gayle McDowell's 'Cracking the Coding Interview' labels the three round types 'Ninja' (coding), 'Pirate' (design), and 'Jedi' (behavioral).

Read the rest of the framework notes

'Jedi' is the one that has clearly stuck in current candidate reporting; treat 'Ninja'/'Pirate' as folklore rather than something you'll hear on your loop. Separately, multiple prep sources (most credibly interviewing.io's hiring-process guide) describe engineer-hiring evaluation mechanics: coding rounds tend to drive a binary hire/no-hire signal, system/product design rounds tend to inform leveling, and behavioral + design feedback carry outsized weight when a panel is split.

These mechanics are consistent across sources but are not an 'official named framework' Meta has published. Treat them as well-corroborated practitioner reporting rather than an Amazon-LP-style artifact. Also worth flagging: Meta separately publishes five broader platform principles (Give people a voice; Build connection and community; Serve everyone; Keep people safe and protect privacy; Promote economic opportunity), but these describe product philosophy, not interview evaluation criteria, and prep sources do not treat them as an interview rubric.

03 / Coach's notes

Coach's tips.

  1. 01

    Treat the technical phone screen as a speed test, not just a correctness test: multiple independent sources report Meta interviewers weight 'did you land on a working algorithm quickly' above 'is the final code bug-free' on the first problem. Narrate your approach out loud and get to a working solution before over-polishing.

  2. 02

    For behavioral ('Jedi') answers, explicitly frame stories around individual ownership and quantified impact rather than team narrative. Prep sources consistently warn that Meta interviewers are trained to probe 'what did YOU specifically do,' and stories that read as 'we' rather than 'I' read weaker even when the outcome was good.

  3. 03

    Since Meta has no single official behavioral rubric, hedge your prep by mapping 2-3 stories to each of the six stated values (Move Fast, Focus on Long-Term Impact, Build Awesome Things, Live in the Future, Be Direct and Respect Your Colleagues, Meta, Metamates, Me) rather than assuming interviewers will name-check them the way Amazon interviewers name Leadership Principles. The mapping is a coaching heuristic, not something Meta asks candidates to do explicitly.

  4. 04

    Clarify with your recruiter which design track you'll get: System Design (distributed systems, scale, storage) versus Product Design/Architecture (API design, client-server flow, product evolution), since prep material and even evaluation emphasis differs meaningfully between the two, and getting this wrong wastes prep time.

  5. 05

    If you're scheduled for the newer AI-assisted coding round, practice reviewing and critically editing AI-generated diffs rather than just writing code from scratch. Reported failure modes are candidates blindly accepting incorrect AI suggestions or being unable to explain, test, or modify AI-proposed changes. If you're E7+ or a Manager-1 candidate, plan for this to be your only coding round, so don't skip design prep to over-rotate on it.

  6. 06

    Understand that passing the onsite loop is not the final gate: team matching happens before the offer and is commonly the slowest, most variable part of the whole process (weeks, not days). Ask your recruiter early about the size of the team-matching pool, and don't assume a fast finish just because the loop went well; Meta's own careers FAQ frames the realistic end-to-end timeline in months, not weeks.

04 / The research

Where this page comes from.

Compiled from public sources across 18 sites.

sites
18
coaching and prep guides
31
candidate reports and forums
2

Source composition

Where the research comes from

Grouped by the kind of site that published each source: Meta's own domain, candidate forums, and everything else.

  • Coaching and prep guides31
  • Candidate reports and forums2

05 / The questions

Reported questions, by category.

Showing questions for

Roles are grouped by the category each question fits, not by the source.

Coding / Data Structures & Algorithms

(phone screen + onsite coding rounds)

8 questions

  1. Two Sum: find indices of two numbers in an array that add up to a target

    Tests hash-map lookup; one pass storing complements gives the O(n) solution.

  2. Merge Intervals: merge all overlapping intervals in a list

    Tests sorting plus a linear sweep; compare each start with the current merged interval's end.

  3. Valid Palindrome / Valid Palindrome II (and the harder, DP-flavored follow-up, Valid Palindrome III, sometimes used to probe further)

    Tests two-pointer string checks; Valid Palindrome III escalates to interval dynamic programming.

  4. Dot Product of Two Sparse Vectors: compute the dot product of two large mostly-zero vectors efficiently, with follow-up discussion of naive vs. hashmap vs. index-value-pair approaches

    Tests exploiting sparsity; compare hashmap vs index-value pairs and the two-pointer merge follow-up.

  5. Random Pick with Weight: pick a random index proportional to its weight, with a follow-up on optimizing for millions of repeated calls

    Tests prefix sums plus binary search; preprocessing makes millions of repeated picks cheap.

  6. Binary tree problems, e.g., Validate Binary Search Tree, Binary Tree Right Side View, Binary Tree Maximum Path Sum, Flatten Binary Tree to Linked List

    Tests tree recursion and traversal; BST validation with bounds, right side view via BFS, and path sums.

  7. Add Strings / Add Two Numbers: add two numbers represented as strings or linked lists without built-in big-integer conversion

    Tests digit-by-digit arithmetic with carry handling, without big-integer shortcuts.

  8. LRU Cache design: implement get/put with O(1) operations

    Classic design; hash map plus doubly linked list gives O(1) get and put.

Reported problems, listed so you know what to expect. Practice them in your own editor or on the platform you prefer; SupaCV's practice mode coaches how you talk through them.

System Design / Product Architecture

(onsite design round)

8 questions

  1. Design the Facebook/Instagram News Feed: content ranking, fan-out strategy, caching, and read-heavy scaling

    System design; content ranking, fan-out on write vs read, caching, and read-heavy scaling.

  2. Design Facebook Messenger or a generic real-time messaging app (WhatsApp-style): delivery/read receipts, multi-device sync, message ordering

    System design; message ordering, delivery and read receipts, and multi-device sync.

  3. Design a search typeahead/autocomplete service: top-k query suggestions as a user types

    System design; trie or prefix index with top-k suggestion ranking under tight latency.

  4. Design a URL shortener

    System design staple; key generation, collision handling, and caching the redirect path.

  5. Design a rate limiter

    System design; token bucket vs sliding window algorithms and distributed enforcement.

  6. Design a web crawler

    System design; frontier management, politeness, deduplication, and distributed coordination.

  7. Design Instagram (photo-sharing feed, comments, likes, notifications at scale)

    System design; media storage, feed generation, and notification fan-out at scale.

  8. For Product Architecture-flavored versions: focus shifts from raw scalability to API design, client-server interaction patterns, and how the product/feature evolves over time and across platforms

Reported problems, listed so you know what to expect. Practice them in your own editor or on the platform you prefer; SupaCV's practice mode coaches how you talk through them.

AI-Assisted Coding Round

(piloted from Q4 2025, expanding through 2026 for E3-E6, and reported as the sole coding round for E7+/Manager-1; verify with your recruiter whether it applies to your loop)

4 questions

  1. Given a multi-file existing codebase (hundreds to thousands of lines) in a CoderPad environment with a file explorer, editor, and AI-assistant panel, use the in-editor AI assistant (model options reportedly include GPT-5, Claude Sonnet/Haiku, Gemini 2.5 Pro, or Llama 4, depending on rollout) to add a requested feature

    Tests navigating an unfamiliar codebase and directing an AI assistant to ship a scoped feature.

  2. Review and decide whether to accept, reject, or modify AI-generated code diffs for correctness and edge cases

    Tests code-review judgment; spotting correctness and edge-case flaws in AI-generated diffs.

  3. Diagnose and fix a bug in an unfamiliar section of the provided codebase with AI assistance

    Tests debugging in unfamiliar code; isolating the fault and validating the fix with AI assistance.

  4. Explain trade-offs of an AI-suggested implementation versus an alternative approach you'd choose manually

    Tests trade-off articulation; comparing an AI-suggested implementation against your own alternative.

Reported problems, listed so you know what to expect. Practice them in your own editor or on the platform you prefer; SupaCV's practice mode coaches how you talk through them.

Every role

Behavioral / 'Jedi' round

(values-fit and past-experience interview)

8 questions

  1. Tell me about a time you had to make a technical decision with incomplete information

  2. Describe a project you led that you're most proud of: what was the measurable impact?

  3. Tell me about a time you fundamentally disagreed with a senior engineer or a peer: how did you resolve it?

  4. Tell me about a time you received harsh or critical feedback (e.g., in a code review): how did you respond?

  5. Tell me about a time you sacrificed your own progress/velocity to unblock a struggling teammate

  6. Describe a serious mistake you made (e.g., one that broke production): what did you learn and change?

  7. Tell me about a time you had to learn a new technology or stack quickly to solve a problem

  8. Why do you want to work at Meta? / What excites you about this role or team?

Your next step

Interviewing at Meta?

Make sure your resume gets you to the first round.

Score your resume, then prep for Meta

Practicing for Meta against your resume is in Interview Prep on Career.

Common questions

Quick answers.

How many stages are in Meta's interview process?

Meta has 8 stages, set out in The process.

What framework does Meta use to evaluate candidates?

The framework behind Meta's evaluation is set out in What Meta scores.

What kinds of questions does Meta ask?

Meta's questions fall into 6 categories, grouped in The questions.

How reliable is this Meta interview playbook?

It is compiled from public sources, and the evidence is in The research.

Sources (33)

Every public source this page was compiled from. Links open in a new tab.

18 sites, by sources

  • igotanoffer.com5
  • hellointerview.com4
  • metacareers.com3
  • prepfully.com3
  • aced.io3
  • interviewing.io2
  • productalliance.com2
  • tryexponent.com1
  • geeksforgeeks.org1
  • prachub.com1
  • leetcode.com1
  • medium.com1
  • designgurus.substack.com1
  • glassdoor.com1
  • fortune.com1
  • linkedin.com1
  • ravi-mehta.com1
  • stellarpeers.com1
  1. 01metacareers.com/culture
  2. 02metacareers.com/hiring-process
  3. 03metacareers.com/swe-prep-onsite
  4. 04interviewing.io/guides/hiring-process/meta-facebook
  5. 05interviewing.io/blog/how-to-use-ai-in-meta-s-ai-assisted-coding-interv…
  6. 06hellointerview.com/blog/meta-ai-enabled-coding
  7. 07igotanoffer.com/blogs/tech/facebook-software-engineer-interview
  8. 08igotanoffer.com/en/advice/meta-phone-screen-interview
  9. 09igotanoffer.com/blogs/tech/facebook-interview-process
  10. 10igotanoffer.com/en/advice/meta-team-matching
  11. 11igotanoffer.com/blogs/product-manager/behavioral-interview-questions-t…
  12. 12hellointerview.com/guides/meta/e5
  13. 13hellointerview.com/guides/meta/e3
  14. 14hellointerview.com/guides/meta/e6
  15. 15prepfully.com/interview-guides/meta-software-engineer
  16. 16prepfully.com/interview-guides/meta-ai-assisted-coding-interview
  17. 17tryexponent.com/blog/meta-system-design-interview
  18. 18geeksforgeeks.org/system-design/meta-facebook-system-design-interview-qu…
  19. 19prachub.com/resources/20-common-meta-behavioral-interview-question…
  20. 20leetcode.com/discuss/interview-question/algorithms/124823/facebook-…
  21. 21medium.com/@johnadjanohoun/metas-most-asked-coding-interview-ques…
  22. 22designgurus.substack.com/p/meta-system-design-interview-prep
  23. 23glassdoor.com/Interview/Meta-Interview-Questions-E40772.htm
  24. 24fortune.com/2022/02/16/mark-zuckerberg-meta-facebook-corporate-val…
  25. 25prepfully.com/interview-guides/meta-product-manager
  26. 26productalliance.com/post/how-to-ace-metas-product-sense-interview-framewor…
  27. 27aced.io/questions/799/how-to-improve-facebook-marketplace
  28. 28linkedin.com/pulse/facebook-pm-interview-samplemock-question-librar…
  29. 29ravi-mehta.com/facebook-product-manager-interview
  30. 30stellarpeers.com/meta-product-manager-interview-process
  31. 31aced.io/blog/meta-product-sense-interview
  32. 32productalliance.com/guides/meta-pm-interview-cheat-sheet
  33. 33aced.io/questions

Related reading

How similar companies interview

Compare Meta's process with other Big Tech companies in this playbook: