FAANG / MAANG+ Coding Interview Questions
Essential Resources
Problem Collections
Company-Specific Questions
AI & Machine Learning
Programming Resources
1:1 Help
What Changed in 2026
The interview format shifted more this past year than in the previous five. The headline: AI-assisted rounds went mainstream, and where AI is allowed the rubric moved to verification: test before you trust the output, and be able to explain it.
- Meta rolled out an AI-enabled coding round (3-panel CoderPad), now reaching SWE and EM roles up through E7/M2. Google is piloting an AI-assisted code-comprehension round with Gemini, while simultaneously bringing back an in-person round to curb cheating. LinkedIn made its AI-enabled round standard. DoorDash publicly rebuilt its interviews around AI.
- OpenAI added an agentic coding round in beta, the only live round where AI is allowed (the take-home also permits it for Applied AI roles). Anthropic runs a split policy (AI permitted on the take-home, banned in most live rounds, though some 2026 onsites now include a pull-request review round with Claude Code allowed) and has redesigned that take-home three times because Claude kept beating it. Sierra dropped algorithm interviews entirely.
- ByteDance and Palantir explicitly ban AI use. Apple reports no AI round at all. Amazon is mixed: some 2026 OAs replace one coding question with a 60-minute AI-assisted repository-debugging task (VS Code-like environment, Spring Boot or MERN app, failing tests, built-in assistant), while other candidates in the same season still get two traditional questions.
- Work trials are the AI-startup norm: Cursor runs paid 8-9 hour onsite projects; OpenAI’s take-home is a paid (~$1,000) work trial.
- Netflix introduced formal engineering levels (E1-E7). The same answer is now scored against your target level.
Full breakdown in FAANG-Recent-Questions.md and the AI Labs guide.
Table of Contents
Company Questions
AI Labs & AI Companies: full guide
| SNo. |
Company |
Signature Round |
| 1. |
Google DeepMind |
2-hour rapid-fire technical quiz (CS + math + stats + ML) |
| 2. |
xAI |
Proctored CodeSignal; concurrency-at-scale extensions |
| 3. |
Mistral AI |
LLM knowledge quiz; ML coding from scratch |
| 4. |
Meta Superintelligence Labs |
AI-enabled coding round |
| 5. |
Amazon AGI |
Transformer debugging + Leadership Principles |
| 6. |
Perplexity AI |
Founder final round; RAG/search depth |
| 7. |
Scale AI |
Card-game OOP simulation; debugging round |
| 8. |
Cohere |
Production infra code (Python/Go), no LeetCode tricks |
| 9. |
Hugging Face |
Open-source take-home (Spaces demo) |
| 10. |
Cursor (Anysphere) |
Paid 8-9 hour onsite project on the real codebase |
| 11. |
Together AI |
CUDA kernel take-home |
| 12. |
Groq |
Compiler/LPU depth (NDA before first interview) |
| 13. |
Cerebras |
Two LC Mediums in 45 minutes |
| 14. |
ElevenLabs |
Product decomposition round |
| 15. |
Waymo |
Modern C++; correctness over speed |
| 16. |
Character.AI |
ML coding + real-time chat design |
| 17. |
Sierra AI |
Plan -> Build (2h with AI) -> Review; no algorithms |
| 18. |
Glean |
2-hour on-the-spot build assignment |
| 19. |
Runway |
Craft deep-dive; GPU pipeline design |
| 20. |
Snowflake |
Data-processing twists; Cortex/AI platform design |
| 21. |
Safe Superintelligence & Thinking Machines Lab |
No verified question bank; network-driven hiring (what is known) |
Quick Start Guide
Beginner Track (0-3 months)
- Start with Blind 75 for fundamentals
- Practice 2-3 problems daily focusing on patterns
- Review Python Resources for clean code
Intermediate Track (1-6 months)
Not sure which track you are on? A single mock interview tells you faster than another week of reading.
Advanced Track (Targeting specific roles)
AI Lab Track (OpenAI, Anthropic, DeepMind, Mistral, xAI)
- Read the AI Labs & AI Companies guide. These loops look nothing like FAANG
- Practice implementing attention, tokenizers, and sampling from scratch: no libraries
- Prepare for debugging rounds (broken Transformers, planted bugs) and progressive multi-level problems
- Expect concurrency follow-ups on ordinary problems, and values/mission rounds that are real gates
Mock Interviews and Mentorship
Reading questions gets you to the door. Doing them under time pressure, with someone watching, gets you through it. If you want a mock interview, a prep plan built around one company’s loop, or a second opinion on where you stand, you can book time with the maintainer of this repo:
Sessions cover FAANG, AI lab and startup loops, including the AI-assisted rounds described in this guide.
FAANG Must Do Problems
View Problems
Google
View 45 Problems (2025-2026 Most Frequent)
2026 changes: an AI-assisted “code comprehension” round is piloting, in which you debug and optimize an existing codebase with Gemini available; interviewers score AI fluency (prompting, output validation, debugging AI output). An in-person round has been reinstated to curb AI-assisted cheating. The Googleyness round is now part-technical. Roughly 19% of reported problems are Hard. Segment tree / BIT problems are a Google-distinctive category.
Custom problems: restaurant waitlist data structure; number of lakes on an island; root an undirected acyclic graph as a binary tree; top-K most talkative users from chat logs; network of teleporters; car rental booking overlap.
View 45 Problems (2025-2026 Most Frequent)
2026 changes: The AI-enabled coding round is rolling out to all SWE roles, 60 min in a 3-panel CoderPad (file explorer, editor, AI chat; GPT-5, Claude Sonnet, Gemini, Llama 4 available; AI reads files but cannot edit). Three phases: fix a bug -> build a 120+ line feature -> optimize for larger datasets. Scored on problem solving, code quality, verification, and communication. For E4-E5 it randomly replaces one of the two coding rounds; at E6 it also replaces one of two, so a traditional CoderPad round normally remains. Behavioral weight increased, it can single-handedly downlevel E5 to E4. Candidates increasingly get variants of tagged problems.
AI-round problems (~9 in rotation): Maze Solver with Path Printing; Maximize Unique Characters from Word List; Card Game (Three Cards Summing to 15); Friend Recommendation System.
Amazon
View 45 Problems (2025-2026 Most Frequent)
2026 changes: HackerRank OA = 2 coding problems (~70 min) + Work Simulation (~20 min) + Work Style Assessment; the SDE II OA adds a 20-min System Design scenario. ~75-80% of OA problems are Medium, wrapped in Amazon-themed framing (servers, warehouses, parcels). Onsite is ~50/50 coding vs Leadership Principles in every round, plus Bar Raiser. Rising: Dijkstra/weighted-shortest-path problems. AI-assisted OA section, mixed reports: since mid-2026 some OAs swap the second coding question for a ~60-min repository-debugging task in a VS Code-like environment (pick a backend stack such as Spring Boot or MERN, make failing tests pass, a built-in assistant is provided; external tools are not). Other candidates in the same cycle still get two traditional questions. Amazon also rotates custom OA sets aggressively, so pattern prep beats memorization.
Custom OA problems (2026): Server Allocation Cost; Warehouse Distribution; Minimum Array Operations; Bug Sorting by Frequency & Code; Distribute Parcels; Suitable Warehouse Locations; Min Cost To Add New Roads (Hard); Dropped Requests (Hard).
Apple
View 30 Problems (2025-2026 Most Frequent)
Still radically team-dependent. No unified loop: some teams ask standard LC mediums, embedded/hardware teams ask C/C++ memory questions, services teams ask API design or debug-broken-code. 2026: design-style coding questions are disproportionately common; loops for experienced hires run 8-9 rounds over several weeks. No AI-assisted rounds reported: human-only interviews graded on correctness, memory behavior, and boundary handling.
Custom problems: memory-efficient ProRAW image decoder (<=1GB RAM); lock-free queue for watchOS sensor data; debug a PyTorch U-Net shape mismatch; Bag-of-Words similarity search; Library Management System OOP design.
Netflix
View 22 Problems (2025-2026 Most Frequent)
Biggest 2026 change: formal engineering levels. Netflix moved from a single “Senior Engineer” rung to an explicit multi-band ladder (~E1/L4-E7). The same coding answer is now scored against your target level, so an answer that passes at E4 can fail at E6 for being “too tactical.” Loops are decentralized and team-owned; the hiring manager is involved from the first screen. Coding favors practical mediums re-skinned with Netflix domain (shows, playlists, watch history). The culture/Keeper Test round is mandatory in every loop.
Custom problems: TTL Cache with LRU eviction; Weighted Eviction Cache; Versioned Key-Value Store; Streaming Word Counter; group users by overlapping last-K watched movies; playlist add/remove/shuffle; rate limiter with graceful degradation when the limiter itself fails.
Microsoft
View 30 Problems (2025-2026 Most Frequent)
2026 changes: Process compressed. OA (2 mediums) then 4 virtual onsite rounds usually on a single day; SDE2 loops = 2-3 DSA + LLD + HLD + hiring manager. The “As Appropriate” (AA) round is formalized and run by Principal EMs; ~85% who reach it get offers, but it retains veto power. Behavioral “growth mindset” scoring is level-banded. AI-assisted rounds are org-specific, not universal: mostly CoreAI/Copilot teams; ask your recruiter.
Custom problems (2026): reconstruct DNA payload strings from tagged fragments; build a CSV query engine; in-memory URL shortener; memory allocation simulation; debug broken queue code; combined rate-limiter + LFU round; traverse an org chart by level; minimum moves on a grid with k-cell jumps.
LLD: Vending Machine, Parking Lot, Snake & Ladder, Elevator Control System (SOLID + State/Strategy probed).
LinkedIn
View 30 Problems (2025-2026 Most Frequent)
2026 changes: The AI-enabled coding round is now standard: one of two coding rounds, on CoderPad with an AI chat panel (Claude/Opus tiers). The AI cannot edit code: you paste and verify. Graded on a 4-point scale where 3 passes, relative to other candidates. The follow-ups are the real bar: after working code, questioning pivots to concurrency/thread safety, scaling, malformed input, and production readiness.
Custom problems: structured data processing from scratch (parse JSON-like objects, AI-round staple); Merge Intervals reframed as a data-structure-choice design problem; LRU Cache with thread-safety and scaling follow-ups; merge two n-ary trees by key rules; count trips from vehicle logs.
OpenAI
View Problems (2025-2026 -- Production-Oriented)
OpenAI interviews focus on practical engineering over LeetCode puzzles. Problems are drawn from a bank of ~8 core challenges with progressive difficulty layers, and the bank churns. Python is strongly recommended.
New in 2026: (1) Agentic coding round (beta): you get an existing codebase and must add features scoped “too large and complex to tackle by hand,” so you’re expected to drive an AI coding agent. This is the only live round where AI is permitted. Every other interview strictly prohibits it (the take-home is the one other carve-out, and only for Applied AI roles). (2) The 48-hour take-home is now a paid work trial (~$1,000) under NDA, graded like a senior engineer’s PR review, “missing test coverage” is the single most-cited rejection reason. (3) Loop = 2 coding + 1 system design + behavioral + hiring manager, plus a 45-min project presentation round.
Core Custom Problems (Most Frequently Reported)
| No. |
Problem |
Difficulty |
Category |
Context |
| 1 |
KV Store Serialize/Deserialize |
Hard |
Design / Strings |
Multi-part: basic serialization, file persistence, multithreading, versioned store |
| 2 |
CD Directory Navigation |
Hard |
String / Path Resolution |
Implement cd() with relative/absolute paths, .., ., ~, symlink cycle detection |
| 3 |
Excel/Spreadsheet Engine |
Hard |
Graph / Design |
getCell() O(1), setCell() with formula dependencies, circular dependency detection |
| 4 |
In-Memory Database |
Hard |
Database Design |
select() with WHERE, AND, ORDER BY, comparison operators – no SQL parsing |
| 5 |
Resumable Iterator |
Hard |
Iterator / State |
Stateful iterator with getState()/setState(); now up to 6 parts: lists -> multi-file -> async -> 2D -> 3D |
| 6 |
Async Node Counting |
Hard |
Distributed / Trees |
Count tree nodes using only async parent-child messaging |
| 7 |
Dependency Version Finder (new 2026) |
Medium-Hard |
Iterative Refinement |
Find earliest version supporting a feature; requirements evolve as test cases are revealed |
| 8 |
GPU Credit Allocation |
Hard |
Design / State |
Half-open intervals [start, expiration), consume soonest-expiring first, balance at any timestamp |
| 9 |
Versioned KV Store |
Hard |
Design |
Auto-versioning; follow-ups: global vs per-key locks, optimistic locking, disk persistence |
| 10 |
Token Consumption Log Parser |
Easy-Medium |
Parsing |
Parse API-call logs, total tokens per user, sort by user ID |
LeetCode-Equivalent Problems
Anthropic
View Interview Guide (2025-2026 -- Custom Problems + Concurrency Round)
Anthropic uses a CodeSignal OA followed by a 4-6 hour onsite with 4-6 rounds. Python expected. Problems are drawn from a bank of ~6 core custom challenges with progressive difficulty layers, and recruiters tell you which prompt family you’ll get days beforehand.
Split AI policy: AI tools are prohibited in most live interviews (candidates have been dropped for using them), but explicitly permitted on the performance take-home. Since mid-2026 some onsites include a round where you review four pull requests with Claude Code available, then improve one. Google/Stack Overflow are allowed in live coding.
The performance take-home has been redesigned three times because Claude kept beating it (Anthropic engineering blog + TechCrunch, Jan 2026): V1 (2024) was a 4-hour simulated-accelerator optimization; V2 (mid-2025) was cut to 2 hours after Claude Opus 4 outperformed most humans; V3 (late 2025) is a Zachtronics-puzzle-style constrained instruction set where you minimize instruction count with no built-in debugging tools, building your own tooling is part of the test. The original is open-sourced at anthropics/original_performance_takehome.
The Values/Culture round (45 min) is the #1 failure point: identical across all roles and levels. It’s NOT behavioral/STAR: it evaluates holding complexity, admitting knowledge gaps, second-order reasoning, and intellectual honesty. Scripted STAR stories are the top failure mode, and measured skepticism about the mission scores better than performed enthusiasm.
Core Custom Coding Problems (Most Frequently Reported)
| No. |
Problem |
Difficulty |
Category |
Context |
| 1 |
In-Memory Database |
Hard |
Design |
4 levels: SET/GET/DELETE -> filtered scans -> TTL -> backup/restore |
| 2 |
Web Crawler |
Hard |
BFS / Concurrency |
BFS crawl -> multithreaded/async optimization |
| 3 |
LRU Cache (Bugfix + Extend) |
Hard |
Design / Debugging |
Fix bugs, add persistence, handle *args/**kwargs |
| 4 |
Stack Trace / Profiler |
Hard |
Parsing / Design |
Convert sampling data to chronological events |
| 5 |
Tokenization Engine |
Hard |
String / NLP |
Greedy longest-match tokenization with unknown-token merging; also a code-review exercise |
| 6 |
Distributed Mode/Median |
Hard |
Distributed Systems |
Compute across 10 nodes with bandwidth constraints |
| 7 |
Record Store |
Hard |
Progressive OA |
In-memory DB + conditional writes + historical (“at timestamp”) queries; L4 = compression/persistence |
| 8 |
Bank Ledger |
Medium-Hard |
Progressive OA |
Account creation -> merging -> delayed cashback -> spending analytics |
| 9 |
Recipe Catalog / Task Tracker |
Easy-Medium |
Progressive OA |
Metadata storage, search by ingredient/prep time; priority + deadline tracking |
| 10 |
Profiler Trace Denoising |
Hard |
Algorithms |
Filter short-lived calls; emit events only after N consecutive appearances |
LeetCode Practice Problems (Mapped to Anthropic’s Focus Areas)
Key Focus Areas: AI safety/alignment (Constitutional AI, RLHF, red-teaming), systems engineering (distributed training, inference optimization), concurrency/parallel programming, Transformer architecture depth. Prep material for the values round: Core Views on AI Safety + the Responsible Scaling Policy.
Palantir
View 25 Problems + Unique Interview Format (2025-2026)
Palantir’s onsite gives you 3 of 4 round types: Decomposition, System Design, Re-engineering (Debugging), and Coding. Each round includes 20 min of behavioral questions. AI use is strictly prohibited in interviews: a notable divergence from the industry’s 2026 drift toward AI-assisted rounds.
The OA is a 3-part practical HackerRank (~90 min): one coding (shape classes OOP) + one SQL (sessions-per-city 3-table join) + one REST API task (paginated restaurant endpoint). Not pure DSA.
Meritocracy Fellowship (launched 2025): an alternative pipeline for high-school grads (SAT >= 1460 / ACT >= 33, $5,400/mo, 4 months). 22 hired from 500+ applicants; successful fellows interview for full-time roles without a degree.
Most Frequently Asked Coding Problems
Unique Interview Rounds
- Decomposition: Break down open-ended real-world problems. Non-coding. AI-flavored prompts (new in 2026): an insurer wants LLM-powered claim summarization; a logistics firm wants an agent to reroute shipments; unify bank fraud detection across legacy systems; retailer demand forecasting; a platform for 500 data sources. Classics: “Design tech to help elderly cook safely”; chess; parking garage; social graph; infection spread; taxi dispatch.
- Re-engineering (Debugging): Debug 500-1000 lines of buggy code with red herrings; sometimes proprietary-library docs are supplied
- System Design: Focus on data integration, ontology design, access control (ABAC); correctness and fault tolerance are first-class constraints
- FDSE vs SWE: Forward-deployed roles emphasize client-facing scenarios; backend roles emphasize scale
Databricks
View 30 Problems (2025-2026 Most Frequent -- includes dedicated concurrency round)
Databricks has a unique dedicated concurrency/multithreading round (1 hour), “most companies wave at the topic; Databricks makes it an entire hour.” OA is 4 problems in 70 minutes on CodeSignal (2 easy, 2 medium), webcam-proctored, scored on a scale up to 850. The onsite is fully virtual in 2026: 2 algorithm rounds + concurrency + system design + behavioral.
Small question pool, deep follow-up variations: the same core problems (SnapshotSet, Lazy Array, House Robber variants, Tic-Tac-Toe) recycle with escalating twists, including “now distribute this with Spark” follow-ups.
README 内容过长，已截取前 64 KB。完整内容请前往 GitHub 仓库查看。
评论