FAANG-Coding-Interview-Questions

A curated List of Coding Questions Asked in FAANG Interviews

  • 默认分支main
  • 大小747 KB
  • 更新5 天前
在 GitHub 打开

README

FAANG / MAANG+ Coding Interview Questions

A curated collection of coding, system design, and ML interview questions from top tech companies.
Continuously updated with 2025-2026 interview questions across 44 companies: FAANG, frontier AI labs, and AI-first startups. Last updated November 2026.

GitHub stars GitHub forks License Last Commit PRs Welcome

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

SNo. Company
1. FAANG Must Do Problems
2. Google
3. Meta (Facebook)
4. Amazon
5. Apple
6. Netflix
7. Microsoft
8. LinkedIn
9. OpenAI
10. Anthropic
11. Palantir
12. Databricks
13. Stripe
14. NVIDIA
15. Uber
16. ByteDance / TikTok
17. Airbnb
18. DoorDash
19. Tesla
20. Flipkart
21. Anduril
22. Figma
23. Ramp

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
No. Problem Difficulty Time Complexity Space Complexity
1 Two Sum Easy O(n) O(n)
2 Longest Substring Without Repeating Characters Medium O(n) O(min(m,n))
3 Longest Palindromic Substring Medium O(n²) O(1)
4 Container With Most Water Medium O(n) O(1)
5 3Sum Medium O(n²) O(1)
6 Remove Nth Node From End of List Medium O(n) O(1)
7 Valid Parentheses Easy O(n) O(n)
8 Merge Two Sorted Lists Easy O(n+m) O(1)
9 Merge k Sorted Lists Hard O(n log k) O(1)
10 Search in Rotated Sorted Array Medium O(log n) O(1)
11 Combination Sum Medium O(2ⁿ) O(n)
12 Rotate Image Medium O(n²) O(1)
13 Group Anagrams Medium O(n k log k) O(n k)
14 Maximum Subarray Medium O(n) O(1)
15 Spiral Matrix Medium O(m×n) O(1)
16 Jump Game Medium O(n) O(1)
17 Merge Intervals Medium O(n log n) O(n)
18 Insert Interval Medium O(n) O(n)
19 Unique Paths Medium O(m×n) O(m×n)
20 Climbing Stairs Easy O(n) O(1)
21 Set Matrix Zeroes Medium O(m×n) O(1)
22 Minimum Window Substring Hard O(n) O(k)
23 Word Search Medium O(m×n×4ᵏ) O(k)
24 Decode Ways Medium O(n) O(n)
25 Validate Binary Search Tree Medium O(n) O(h)
26 Same Tree Easy O(n) O(h)
27 Binary Tree Level Order Traversal Medium O(n) O(n)
28 Maximum Depth of Binary Tree Easy O(n) O(h)
29 Construct Binary Tree from Preorder and Inorder Traversal Medium O(n) O(n)
30 Best Time to Buy and Sell Stock Easy O(n) O(1)

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.

No. Problem Difficulty Category
1 Two Sum Easy Hash Map / Arrays
2 Number of Islands Medium Graph / DFS
3 Merge Intervals Medium Intervals / Sorting
4 LRU Cache Medium Design / Linked List
5 Validate Binary Search Tree Medium Trees / BST
6 Course Schedule II Medium Graph / Topological Sort
7 Longest Substring Without Repeating Characters Medium Sliding Window
8 Trapping Rain Water Hard Two Pointers / Stack
9 Serialize and Deserialize Binary Tree Hard Trees / Design
10 Kth Largest Element in an Array Medium Heap / Quickselect
11 Median of Two Sorted Arrays Hard Binary Search
12 Group Anagrams Medium Hashing / Strings
13 Word Ladder Hard BFS / Graphs
14 Merge K Sorted Lists Hard Heap / Linked List
15 Container With Most Water Medium Two Pointers
16 Top K Frequent Elements Medium Heap / Hashing
17 Coin Change Medium Dynamic Programming
18 Search in Rotated Sorted Array Medium Binary Search
19 Product of Array Except Self Medium Arrays / Prefix
20 Binary Tree Level Order Traversal Medium Trees / BFS
21 Combination Sum Medium Backtracking
22 Edit Distance Medium Dynamic Programming
23 Minimum Window Substring Hard Sliding Window
24 Implement Trie (Prefix Tree) Medium Trie
25 Accounts Merge Medium Union-Find
26 Sliding Window Maximum Hard Monotonic Deque
27 The Skyline Problem Hard Heap / Divide and Conquer
28 Largest Rectangle in Histogram Hard Monotonic Stack
29 Word Search II Hard Trie / Backtracking
30 Rotting Oranges Medium Graph / BFS
31 Critical Connections in a Network Hard Graph / Tarjan’s
32 Pacific Atlantic Water Flow Medium Graph / Multi-source BFS
33 Network Delay Time Medium Graph / Dijkstra
34 Daily Temperatures Medium Monotonic Stack
35 Find the Safest Path in a Grid Medium BFS / Binary Search
36 The Earliest Moment When Everyone Become Friends Medium Union-Find
37 Checking Existence of Edge Length Limited Paths Hard Union-Find / Offline Queries
38 Longest String Chain Medium DP / Hash
39 Maximum Points You Can Obtain from Cards Medium Sliding Window
40 Step-By-Step Directions From a Binary Tree Node to Another Medium Tree / LCA
41 Swim in Rising Water Hard Binary Search + BFS
42 Detect Squares Medium Design / Geometry
43 Amount of New Area Painted Each Day Hard Sweep Line / Segment Tree
44 Range Sum Query - Mutable Medium Segment Tree / BIT
45 Best Meeting Point Hard Math / Median

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.

Meta (Facebook)

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.

No. Problem Difficulty Category
1 Minimum Remove to Make Valid Parentheses Medium Stack / String
2 Binary Tree Vertical Order Traversal Medium Tree / BFS
3 Basic Calculator II Medium Stack / Expression Parsing
4 Valid Palindrome II Easy Two Pointers / String
5 Kth Largest Element in an Array Medium Heap / Divide & Conquer
6 Lowest Common Ancestor of a Binary Tree Medium Tree / DFS
7 Random Pick with Weight Medium Binary Search / Prefix Sum
8 Subarray Sum Equals K Medium Array / HashMap / Prefix Sum
9 Valid Palindrome Easy Two Pointers / String
10 Two Sum Easy Array / HashMap
11 Binary Tree Right Side View Medium Tree / BFS
12 Top K Frequent Elements Medium Heap / HashMap
13 Merge Intervals Medium Array / Sorting
14 LRU Cache Medium Design / HashMap / Linked List
15 Clone Graph Medium Graph / BFS / DFS
16 Merge k Sorted Lists Hard Linked List / Heap
17 Maximum Swap Medium Math / Greedy
18 Number of Islands Medium Graph / DFS / BFS
19 Accounts Merge Medium Union-Find / Graph
20 Diameter of Binary Tree Easy Tree / DFS
21 Product of Array Except Self Medium Array / Prefix Sum
22 Word Break Medium Dynamic Programming
23 Copy List with Random Pointer Medium Linked List / Hash Table
24 Making a Large Island Hard Graph / DFS / Union-Find
25 Expression Add Operators Hard Backtracking / Math
26 Valid Word Abbreviation Easy String / Parsing
27 Lowest Common Ancestor of a Binary Tree III Medium Tree / Hash Table
28 Convert BST to Sorted Doubly Linked List Medium Tree / Linked List
29 K Closest Points to Origin Medium Heap / Math
30 Interval List Intersections Medium Two Pointers / Intervals
31 Simplify Path Medium Stack / String
32 Insert Delete GetRandom O(1) Medium Design / Hash Table
33 Sliding Window Maximum Hard Monotonic Deque
34 Regular Expression Matching Hard DP / Recursion
35 All Nodes Distance K in Binary Tree Medium Tree / BFS
36 Validate IP Address Medium String Parsing
37 Remove All Adjacent Duplicates in String II Medium Stack
38 Greatest Common Divisor of Strings Easy String / Math
39 Find the Length of the Longest Common Prefix Medium Trie / Hash
40 Toeplitz Matrix Easy Matrix
41 Range Sum Query 2D - Immutable Medium Prefix Sums
42 Diagonal Traverse II Medium Array / BFS
43 Wildcard Matching Hard DP
44 Integer to English Words Hard String / Recursion
45 First Missing Positive Hard Array / Index Cycle

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.

No. Problem Difficulty Category
1 Two Sum Easy Array / Hash Table
2 Number of Islands Medium Graph / DFS / BFS
3 LRU Cache Medium Design / Hash + Linked List
4 Merge Intervals Medium Array / Sorting
5 Group Anagrams Medium Hashing / Strings
6 Top K Frequent Elements Medium Heap / Hashing
7 Task Scheduler Medium Heap / Greedy
8 Trapping Rain Water Hard Two Pointers / Stack
9 Product of Array Except Self Medium Array / Prefix Sum
10 Longest Substring Without Repeating Characters Medium Sliding Window
11 Minimum Window Substring Hard Sliding Window
12 Kth Largest Element in an Array Medium Heap / Quickselect
13 Merge k Sorted Lists Hard Heap / Linked List
14 Course Schedule Medium Graph / Topological Sort
15 Rotting Oranges Medium Graph / BFS
16 Serialize and Deserialize Binary Tree Hard Trees / Design
17 Lowest Common Ancestor of a Binary Tree Medium Trees / DFS
18 Coin Change Medium Dynamic Programming
19 Best Time to Buy and Sell Stock Easy Array / DP
20 Sliding Window Maximum Hard Deque / Sliding Window
21 Reorganize String Medium Heap / Greedy
22 Insert Delete GetRandom O(1) Medium Design / Hash Table
23 Decode String Medium Stack / Strings
24 Koko Eating Bananas Medium Binary Search
25 Largest Rectangle in Histogram Hard Monotonic Stack
26 Car Pooling Medium Array / Prefix Sum
27 Meeting Rooms III Hard Heap / Sorting
28 K Closest Points to Origin Medium Heap / Math
29 Valid Parentheses Easy Stack / String
30 Word Search Medium Backtracking / Matrix
31 Course Schedule II Medium Graph / Topological Sort
32 Design Hit Counter Medium Design / Queue
33 Min Stack Medium Stack / Design
34 Maximum Frequency After Subarray Operation Medium Array / Greedy-DP
35 Maximize Y-Sum by Picking a Triplet of Distinct X-Values Medium Heap / Greedy
36 Max Difference You Can Get From Changing an Integer Medium Greedy / Digits
37 Analyze User Website Visit Pattern Medium Hash / Sorting
38 Minimum Cost to Connect Sticks Medium Heap / Greedy
39 Word Search II Hard Trie + Backtracking
40 Alien Dictionary Hard Topological Sort
41 The Skyline Problem Hard Sweep Line / Heap
42 Burst Balloons Hard Interval DP
43 Count of Smaller Numbers After Self Hard Merge Sort / BIT
44 Max Points on a Line Hard Geometry / Hash
45 Reorder List Medium Linked List

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.

No. Problem Difficulty Category
1 Two Sum Easy Array / Hash Table
2 LRU Cache Medium Design / Hash + Linked List
3 Number of Islands Medium Graph / DFS / BFS
4 Reverse Linked List Easy Linked List
5 Group Anagrams Medium String / Hash Table
6 Valid Parentheses Easy Stack / String
7 Merge Intervals Medium Array / Sorting
8 Word Break Medium Dynamic Programming
9 Product of Array Except Self Medium Array / Prefix Sum
10 Best Time to Buy and Sell Stock Easy Array / DP
11 3Sum Medium Array / Two Pointers
12 Trapping Rain Water Hard Two Pointers / Stack
13 Top K Frequent Elements Medium Heap / Hash Table
14 Course Schedule Medium Graph / Topological Sort
15 Lowest Common Ancestor of a Binary Tree Medium Tree / DFS
16 Serialize and Deserialize Binary Tree Hard Tree / Design
17 Longest Substring Without Repeating Characters Medium Sliding Window
18 Median of Two Sorted Arrays Hard Binary Search
19 Maximum Profit in Job Scheduling Hard Binary Search / DP
20 Bus Routes Hard Graph / BFS
21 Sum Root to Leaf Numbers Medium Trees
22 Check Completeness of a Binary Tree Medium Trees / BFS
23 Time Based Key-Value Store Medium Design / Binary Search
24 Design Hit Counter Medium Design
25 Design Add and Search Words Data Structure Medium Trie / Design
26 Vertical Order Traversal of a Binary Tree Hard Trees
27 Binary Search Tree Iterator Medium Design / Trees
28 Subarray Sum Equals K Medium Prefix Sum
29 Minimum Unique Word Abbreviation Hard Backtracking / Bitmask
30 H-Index Medium Sorting

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.

No. Problem Difficulty Category
1 LRU Cache Medium Design / Hash + Linked List
2 Merge Intervals Medium Array / Sorting
3 Course Schedule II Medium Graph / Topological Sort
4 Top K Frequent Elements Medium Heap / Hash Table
5 Network Delay Time Medium Graph / Dijkstra
6 Daily Temperatures Medium Monotonic Stack
7 Number of Islands Medium Graph / DFS / BFS
8 Serialize and Deserialize Binary Tree Hard Tree / Design
9 Find Median from Data Stream Hard Heap / Design
10 Trapping Rain Water Hard Two Pointers / Stack
11 Edit Distance Medium Dynamic Programming
12 Minimum Window Substring Hard Sliding Window
13 Meeting Rooms II Medium Intervals / Heap
14 Koko Eating Bananas Medium Binary Search
15 Implement Trie (Prefix Tree) Medium Trie / Design
16 Rotating the Box Medium Matrix / Simulation
17 Time Based Key-Value Store Medium Design / Binary Search
18 Logger Rate Limiter Easy Design
19 Word Search II Hard Trie / Backtracking
20 Reconstruct Itinerary Hard Graph / Euler Path
21 Alien Dictionary Hard Topological Sort
22 Parallel Courses Medium Topological Sort

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.

No. Problem Difficulty Category
1 Two Sum Easy Array / Hash Table
2 LRU Cache Medium Design / Hash + Linked List
3 Longest Substring Without Repeating Characters Medium Sliding Window
4 Add Two Numbers Medium Linked List / Math
5 Number of Islands Medium Graph / DFS / BFS
6 Maximum Subarray Medium Array / DP
7 Merge Two Sorted Lists Easy Linked List
8 Copy List with Random Pointer Medium Linked List / Hash Table
9 Rotate Image Medium Array / Matrix
10 Set Matrix Zeroes Medium Array / Matrix
11 Clone Graph Medium Graph / DFS / BFS
12 Binary Tree Level Order Traversal Medium Tree / BFS
13 Median of Two Sorted Arrays Hard Binary Search
14 Regular Expression Matching Hard DP / Recursion
15 Reverse Nodes in k-Group Hard Linked List
16 Asteroid Collision Medium Stack
17 Cheapest Flights Within K Stops Medium Graph / DP / BFS
18 Construct Binary Tree from Preorder and Inorder Traversal Medium Tree / Divide and Conquer
19 Largest Rectangle in Histogram Hard Monotonic Stack
20 Sudoku Solver Hard Backtracking / Matrix
21 Trapping Rain Water Hard Two Pointers
22 Container With Most Water Medium Two Pointers
23 Longest Palindromic Substring Medium DP / Strings
24 Next Permutation Medium Arrays
25 Search in Rotated Sorted Array Medium Binary Search
26 Spiral Matrix Medium Matrix
27 Word Search Medium Backtracking
28 First Missing Positive Hard Arrays
29 Jump Game II Medium Greedy
30 Generate Parentheses Medium Backtracking

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.

No. Problem Difficulty Category
1 Nested List Weight Sum Medium DFS / Recursion
2 Nested List Weight Sum II Medium Stack / DFS
3 Can Place Flowers Easy Array / Greedy
4 Find Leaves of Binary Tree Medium Tree / DFS
5 Max Stack Hard Stack / Linked List / Design
6 All O’one Data Structure Hard Hash Table / Linked List / Design
7 Shortest Word Distance II Medium Hash Table / Design
8 Find the Celebrity Medium Two Pointers / Graph
9 Maximum Subarray Medium Array / DP
10 Maximum Product Subarray Medium Array / DP
11 Merge Intervals Medium Array / Sorting
12 Edit Distance Medium Dynamic Programming
13 Design Add and Search Words Data Structure Medium Trie / Design
14 Word Ladder Hard BFS / Graphs
15 Insert Delete GetRandom O(1) Medium Design / Hash Table
16 Valid Parentheses Easy Stack / String
17 Isomorphic Strings Easy String / Hash Table
18 Decode Ways Medium Dynamic Programming
19 House Robber II Medium Dynamic Programming
20 Combination Sum II Medium Backtracking
21 LFU Cache Hard Design (confirmed in AI-enabled round)
22 Insert Delete GetRandom O(1) - Duplicates allowed Hard Design
23 Max Consecutive Ones III Medium Sliding Window
24 Max Consecutive Ones II Medium Sliding Window
25 Design Authentication Manager Medium Design / TTL
26 Serialize and Deserialize BST Medium Trees / Design
27 Paint House III Hard DP
28 Allocate Mailboxes Hard DP
29 Generate Random Point in a Circle Medium Math / Randomized
30 Valid Perfect Square Easy Binary Search

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

No. Problem Difficulty Category Context
1 LRU Cache Medium Design Inference KV cache – most frequently reported
2 Time Based Key-Value Store Medium Design / Binary Search Model checkpoint storage
3 Snapshot Array Medium Design / Binary Search Model state checkpointing
4 Alien Dictionary Hard Graph / Topological Sort Tokenizer ordering
5 Web Crawler Multithreaded Medium Concurrency / BFS Training data crawling
6 LFU Cache Hard Design Advanced caching
7 Decode String Medium Stack / Strings String processing
8 Word Ladder Hard BFS / Graphs NLP transformations
9 Design Memory Allocator Medium Design / Simulation GPU memory management
10 Game of Life Medium Simulation / Matrix Extended to infinite board
11 Meeting Rooms II Medium Intervals / Heap Interval scheduling
12 Serialize and Deserialize Binary Tree Hard Trees / Design Data persistence
13 Top K Frequent Elements Medium Heap / Hash ML preprocessing
14 Course Schedule II Medium Graph / Topological Sort Dependency resolution

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)

No. Problem Difficulty Category
1 LRU Cache Medium Design
2 Web Crawler Multithreaded Medium Concurrency / BFS
3 Implement Trie (Prefix Tree) Medium Trie / NLP
4 Word Break Medium DP / Strings
5 Design Hit Counter Medium Design
6 Time Based Key-Value Store Medium Design / Binary Search
7 Serialize and Deserialize Binary Tree Hard Trees / Design
8 Merge k Sorted Lists Hard Heap / Distributed
9 Course Schedule II Medium Graph / Topological Sort
10 Number of Islands Medium Graph / DFS
11 Count of Smaller Numbers After Self Hard Merge Sort / BIT (phone screen; O(n log n) required)

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

No. Problem Difficulty Category
1 Merge Intervals Medium Intervals / Sorting
2 Number of Islands Medium Graph / DFS / BFS
3 LRU Cache Medium Design / Hash Map
4 Course Schedule Medium Graph / Topological Sort
5 Course Schedule II Medium Graph / Topological Sort
6 All Ancestors of a Node in DAG Medium Graph / DFS
7 Merge k Sorted Lists Hard Heap / Linked List
8 Trapping Rain Water Hard Two Pointers / Stack
9 Regular Expression Matching Hard DP / String
10 Subdomain Visit Count Medium Hash Map / String
11 Find the Celebrity Medium Array / Logic
12 UTF-8 Validation Medium Bit Manipulation
13 Container With Most Water Medium Two Pointers
14 Max Area of Island Medium Graph / DFS
15 Rotate Image Medium Array / Matrix
16 Integer to English Words Hard String / Math
17 Shortest Path to Get All Keys Hard BFS / Bitmask
18 Construct Quad Tree Medium Tree / Recursion
19 Inorder Successor in BST Medium Tree / BST
20 Contains Duplicate III Hard BST / Bucket Sort
21 Minimum Time Difference Medium String / Sorting
22 Flood Fill Easy BFS / DFS
23 Cheapest Flights Within K Stops Medium Shortest Path
24 Word Ladder Hard BFS / State-Space Search
25 Best Time to Buy and Sell Stock Easy Array

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.

No. Problem Difficulty Category
1 Capacity To Ship Packages Within D Days Medium Binary Search
2 Trapping Rain Water Hard Two Pointers / Stack
3 Max Stack Hard Stack / Linked List / Design
4 All O’one Data Structure Hard Hash Table / Design
5 Word Break Medium DP / Trie
6 Rotting Oranges Medium BFS / Matrix
7 All Nodes Distance K in Binary Tree Medium Tree / BFS / DFS
8 Decode String Medium Stack / Recursion
9 K Closest Points to Origin Medium Heap / Math
10 Asteroid Collision Medium Stack
11 Design Hit Counter Medium Design / Queue
12 Time Based Key-Value Store Medium Design / Binary Search
13 [Snapshot Array](https://leetcode.c

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

暂无评论