Google LeetCode questions& coding interview guide
Google's coding interviews are famous for rewarding clear thinking over memorized tricks. Interviewers care as much about how you reason toward a solution as the solution itself: how you clarify the problem, choose a data structure, reason about complexity, and handle edge cases out loud. This guide breaks down the topics and question patterns Google leans on, and how to practice them so you can pass on your own.
The Google interview format
Expect an initial phone/virtual screen with one or two coding problems in a shared editor, then an on-site (or virtual on-site) loop of roughly four to five 45-minute rounds. Most are data-structures-and-algorithms coding rounds; senior candidates also get a system design round. You'll code in a plain editor with no autocomplete, so fluency in your chosen language matters.
Difficulty: Predominantly LeetCode Medium, with a Hard problem showing up in the later loop rounds. Depth of follow-ups matters more than raw problem count.
Topics Google emphasizes
These are the data-structure and algorithm areas worth prioritizing before a Google coding interview.
- Arrays, strings, and hash maps
- Two pointers and sliding window
- Binary search (including on the answer)
- Trees, tries, and recursion
- Graphs (BFS/DFS, topological sort)
- Dynamic programming
- Heaps and greedy algorithms
- Time and space complexity analysis
Question patterns to drill
We don't republish Google's actual questions — that's not how you get good. Instead, drill these recurring patterns until you can recognize and solve them on your own.
- Transform-a-string / interval-merging style problems that reward a clean two-pointer or sliding-window pass
- Graph traversal framed as a real-world grid, dependency, or network problem
- Tree problems that need recursion plus a follow-up about doing it iteratively
- Optimization problems where the naive answer is obvious and the interviewer pushes you to a better complexity
- Design-lite problems (build a data structure with specific operations in given time bounds)
How to prepare for Google specifically
- Narrate your reasoning — Google grades communication and problem-solving, not just a passing solution.
- Always state brute force first, give its complexity, then optimize; skipping this reads as guessing.
- Ask clarifying questions about input size and constraints before you write code.
- Practice writing bug-free code without an IDE; the editor has no autocomplete or run button in most rounds.
A Google prep plan
- 1Weeks 1–2: Rebuild fundamentals — arrays, hash maps, two pointers, and binary search until they're automatic.
- 2Weeks 3–4: Trees, graphs, and recursion; practice explaining each solution out loud as if to an interviewer.
- 3Weeks 5–6: Dynamic programming and greedy patterns, plus timed mock interviews to build composure.
- 4Final week: Full mock loops that mimic the 45-minute format, focusing on communication and edge cases.
Prefer to learn by doing? Practice with an AI LeetCode tutor →
Google coding interview FAQs
There's no magic number — depth beats volume. Most successful candidates work through a few hundred Medium problems organized by pattern (two pointers, graphs, DP) rather than grinding at random. Practicing with feedback on your reasoning, like an AI tutor that asks follow-ups, gets you further than raw count.
Mostly LeetCode Medium, with a Hard appearing in later rounds. What raises the bar is the follow-ups: interviewers push you to improve complexity, handle edge cases, and sometimes re-implement iteratively.
No. Live interviews are your own work, and using an AI assistant during them can end your candidacy. Use an AI LeetCode tutor to prepare beforehand so you can solve and explain problems unaided.
Other company guides
Practice Google-style problems with an AI tutor
Solve real problems in 8 languages against a live judge, then run AI mock interviews that coach you through the patterns Google tests — so you pass on your own.
Start practicing free