Search the siteSearch in

Interview paths
Question bank
Coding practice
Study materials
Infrastructure lab

Community

Weekly sessions
Peer practice
Contribute

News

Jobs

Open roles
Market intelligence
Companies
Roles

Feedback

Product feedback
Developer feedback

Checking account…

Home
Question bank

Frontier AI technical interview question bank.

Practice coding, system design, behavioral, and project questions for frontier AI roles.

Company

Loading the question catalog…

Preparing this page…

Your current navigation and account state stay available while the next page loads.

Frontier Rehearsal · Independent practice · Not affiliated with OpenAI, Anthropic, or xAI.

Home Interview questions Company interview processes Company reports Role reports Market intelligence Open roles Python lab Materials How it works About Feedback Uploads Feature requests Terms Privacy 中文 English

Home
Question bank

Frontier AI technical interview question bank.

Practice coding, system design, behavioral, and project questions for frontier AI roles.

Company

FREE PRACTICE

Public demos, always pinned.

Start with either demo. They remain here when you filter or sort.

Free account: sign in to practice the marked Free questions.

Newest published questions; deterministic, never randomized.

Demo 1Number search with delayed feedback
Demo 2Design real-time device monitoring and control

Question bank filters

Reset filters

Question groups

Filter question bank by interview groupAll 269 Coding 162 System design 48 Behavioral 29 Hiring manager 16 Project deep dive 14

Topics

Filter question bank by topic tagsAll 269 Python 161 Algorithms 40 Data structures 46 Concurrency 27 Distributed systems 7 Storage & state 43 Networking 7 Reliability 30 ML / AI 27 Data / SQL 24 Security 7 Product design 25 Communication 66

Independent and not affiliated with the companies represented. Collected material does not predict future interview questions.

245 questions

Sort question bankSortFree first Newest Oldest AI ratio ↓AI ratio ↑

Question bank results

Every question has a public brief with the task, constraints, and follow-ups. Labels describe interactive practice access.

Free practice

Topic tags
Communication

xAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Project walkthrough and motivation

Prepare one project story and one reason for the role. The project story is five to eight minutes. Open on a constraint and the choice you made.

Practice focusRequirements - State why you chose that approach, what you turned down, and what you measured. - Lead with the technical choice, not a business result. - Be ready for a redirect to recent work, a deeper bottleneck, or an option you did not use. - The motivation names a real direction you can contribute to, tied to a skill you have already used. - Keep two depths ready: the system shape, and the kernel or model detail. Pick the depth the conversation asks for. - Keep a short reset so a redirect does not lose the thread. Review checklist 1.…

Free practice

Topic tags
Storage & state
Communication

AnthropicBehavioralHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Decision under uncertainty

Tell one story about a consequential decision you made with incomplete evidence. State what was known, what was unknown, the reversible step you chose, and the signal that would make you change course.

Practice focusPreparation - Write the user or stakeholder, the constraint, and the decision in one sentence each. - Choose one measurement or observable fact that can support the answer. - Name one alternative and the reason it was not selected. Review checklist 1. Did I state the concrete constraint before describing the solution? 2. Did I separate my own decision and work from the team's work? 3. Did I name evidence or a measurement rather than relying on confidence? 4. Can I answer every follow-up without changing the original facts? Staged follow-ups 1.…

Free practice

Topic tags
ML / AI
Communication

AnthropicHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Safety and speed trade-off

You own a useful model feature with a fixed launch date and one unresolved safety concern. Explain the concern, bound the plausible harm, choose what ships, and define the evidence required before expanding acc…

Free practice

Topic tags
Reliability
Communication

AnthropicProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Reliability and recovery project

Walk through a system you made more reliable after a serious failure. Cover the failure boundary, your diagnosis, the design change you owned, the rollout, and the evidence that recovery improved.

Free practice

Topic tags
Storage & state
Product design
Communication

OpenAIBehavioralHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Cross-functional disagreement

Tell one story about a disagreement with a product, research, policy, or operations partner. State the shared goal, the competing constraints, what you changed in your own position, and the decision the group m…

Free practice

Topic tags
Storage & state
Reliability
ML / AI
Product design
Communication

OpenAIHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Product judgment under ambiguity

A team can improve model quality, reduce latency, or make failures easier to understand, but can fund only one this quarter. Choose a direction, state the users and evidence behind it, and define what would inv…

Free practice

Topic tags
Storage & state
Communication

OpenAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Cost and latency project

Walk through a project where you reduced latency or cost without weakening correctness. Cover the baseline, bottleneck, alternatives, the change you owned, rollout safeguards, and measured result.

Free practice

Topic tags
Reliability
Product design

OpenAISystem designHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Tool execution platform

Design a platform where an AI assistant invokes approved tools on behalf of users. Include tool registration, argument validation, identity and consent, isolated execution, timeouts, idempotency, audit logs, an…

Free practice

Topic tags
Storage & state
Security
Communication

xAIHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

High-ownership role fit

Explain why a high-ownership role fits you using one project where you found an unowned problem, chose a direction, and delivered a measurable result. State the boundary of your authority and when you asked for…

Free practice

Topic tags
Storage & state
ML / AI
Communication

xAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Scaling bottleneck project

Walk through a project where scale exposed a bottleneck. Cover the workload, the measured limit, your model of the bottleneck, the intervention you owned, and the evidence that the limit moved.

Free practice

Topic tags
Data structures
ML / AI
Communication

xAISystem designHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Real-time inference gateway

Design a gateway for low-latency model inference across several model versions and accelerator pools. Include admission control, routing, batching, deadlines, retries, overload behavior, observability, and safe…

Free practice

Topic tags
Python

AnthropicCodingHalf AI-generated

Python codingWebAssembly + public tests

Reported interview periodNot available

Build an Unbiased Randomness Adapter

Turn an unknown biased bit source into balanced choices, a sampled load estimate, and isolated concurrent batches.

Practice focusYou receive a hardware source raw bit() that independently returns 0 or 1 with one stable but unknown probability of returning 1 . The probability is strictly between zero and one, and no other random source is available. Implement the following four functions. Requirements 1. Extract balanced bits Read the raw source in fresh pairs. Discard pairs whose values match. For the first pair whose values differ, return its first value. A scripted source raises EOFError when exhausted; let that exception propagate, including for a partial pair.…

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

First position of a target in a sorted array

Given a nondecreasing array and a target value, report where the target first appears, or -1 when it is missing. A logarithmic search is expected, even though the same value may occur many times.

Practice focusRequirements - Line one is the element count n . Line two holds the n space-separated integers in nondecreasing order, and is empty when n is 0. Line three is the target. - Return the smallest zero-based index holding the target, as a decimal string with no padding. - Return -1 when the array is empty or the target never appears. - Repeated values are expected, so an ordinary equality hit is not enough; after a match keep searching to the left while remembering the best index seen. - Keep the running time logarithmic in n .…

Free practice

Topic tags
Python
Concurrency

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Per-function exclusive runtime from a call log

A single thread records start and end events for nested function calls. For each function id, work out how many time units it actually spent running, excluding the units spent inside calls it made.

Practice focusRequirements - Line one is the number of functions; ids run from 0 to count - 1 and every id appears at least once. Line two is the number of event lines. - Each remaining line is <id :<start|end :<timestamp with integer fields. Events are chronological and properly nested on one thread. - An end timestamp is inclusive: the unit at that timestamp belongs to the call that is ending, so it contributes timestamp - previous + 1 units. - When a call begins on top of a running call, credit the units since the previous boundary to the caller.…

Free practice

Topic tags
Python
Data structures

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Maximum haul from non-adjacent houses

Each house on a street holds a known amount of money, and robbing two neighboring houses on the same night trips the alarm. Choose houses with no two adjacent to collect the largest possible total.

Practice focusRequirements - Line one is the house count. Line two lists that many non-negative integer amounts, separated by spaces. - Adjacency is positional, so choosing the house at index i rules out only i - 1 and i + 1 . - Return the best total as a decimal string with no separators. - A single house contributes its own amount, and a zero amount is a legal choice, so a street of zeros answers 0 . - Every house is a candidate; skipping a house is never required just because it is small. - run case passes case["stdin"] to solve .…

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Values unique to the first list

Two Python-style lists arrive as literal assignments. Collect every value that belongs to the first list but not the second, keeping the first list's order.

Practice focusRequirements - The input holds two assignment statements, one per entry: list1 = [...] followed by list2 = [...] . Both lists contain integers and may be empty. - The statements may be separated by a real newline or by a literal two-character \n escape; split on either form. - Return the kept values in their original list1 order. - Membership in list2 is decided by value, not by count or position, so a single occurrence there removes every matching occurrence in list1 .…

Free practice

Topic tags
Python
Algorithms

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Longest run of distinct characters

Given a string, measure the longest contiguous stretch in which no character repeats. Report only the length of that stretch.

Practice focusRequirements - The input is a single string, and the trailing newline is not part of the payload. - A stretch must be contiguous, and every character inside it must be distinct. - Return the length as a decimal string, not the slice itself and not its starting position. - Characters compare exactly, so an uppercase and a lowercase letter are different characters. - A string whose characters all match answers 1 , and the longest stretch may sit anywhere in the payload.…

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Peak simultaneous bookings

Given a list of half-open time ranges, report the largest number of ranges that are active at the same instant.

Practice focusRequirements - Line 1 holds the number of bookings. Each following line holds start end , two integers with start strictly less than end . - Treat a booking as occupying the half-open interval [start, end) . - Because the interval is half-open, a booking that ends at a time does not conflict with one that starts at that same time. - Return the peak number of concurrent bookings as a decimal string. - An empty booking list returns "0" . - run case(case: dict) forwards case["stdin"] to solve .…

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Merge three ordered lists without repeats

Combine three nondecreasing integer lists into one nondecreasing list in which every distinct value appears once, then print it as a single space-separated line.

Practice focusRequirements - The input has exactly three lines. Each line starts with a count k followed by k nondecreasing integers. - The merged result must stay sorted and must contain each distinct value exactly once. - Values may be negative, zero, or positive. Duplicates can appear within one line and across lines. - Print the distinct values in one line separated by single spaces, with no leading or trailing space. - If all three lists are empty, return an empty string. - run case(case: dict) forwards case["stdin"] to solve .…

Free practice

Topic tags
Python
Data structures

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Next arrangement in lexicographic order

Rearrange an integer list into the next larger lexicographic arrangement, wrapping around to the smallest arrangement when no larger one exists.

Practice focusRequirements - Line 1 holds the length n . Line 2 holds n integers separated by spaces. - Compare arrangements position by position from the left. The next arrangement is the smallest arrangement strictly greater than the current one. - Duplicate values are allowed. - If the current arrangement is already the largest, return the smallest arrangement, with values in ascending order. - Print the resulting list in one space-separated line, keeping the same length and the same multiset of values. - run case(case: dict) forwards case["stdin"] to solve .…

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Count distinct values in an ordered list

Given a nondecreasing integer sequence, report how many different values it contains.

Practice focusRequirements - The input is one line of space-separated integers in nondecreasing order. - Equal values are therefore adjacent, and the sequence may contain negative values. - Return the number of distinct values as a decimal string. - You do not need to print or return the compacted sequence itself. - run case(case: dict) forwards case["stdin"] to solve . - Do not append a trailing newline to the returned string; return the value exactly once.

Free practice

Topic tags
Python
Algorithms

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Zero rows and columns around each zero

Given a matrix, set every cell in a row or column that contains a zero to zero, deciding which rows and columns to clear from the original values rather than from the partially rewritten matrix.

Practice focusRequirements - Line 1 is rows cols . The next rows lines each hold cols integers. - For every cell that is 0 in the original matrix, its entire row and entire column become 0 . - Decide all affected rows and columns before writing any zeros, so zeros created by the update must not widen the cleared area. - Return the matrix with rows in their original order, values separated by single spaces and rows separated by newlines, with no extra text. - A matrix with no zeros is returned unchanged. - run case(case: dict) forwards case["stdin"] to solve .…

Free practice

Topic tags
Python
Storage & state

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Normalize an absolute filesystem path

Read one absolute path and return its canonical form: empty segments and . segments disappear, .. folds into the parent, and exactly one slash separates the surviving directories.

Practice focusRequirements - The input is a single absolute path; it may repeat separators and may end with a slash. - A whole segment equal to . is dropped; only a whole segment equal to .. means "go up". - .. removes the most recently kept segment; when nothing is left to remove it has no effect, so climbing past the root stays at the root. - Join the remaining segments with one leading slash, no doubled slashes, and no trailing slash. - Return / alone when no segment survives.…

Free practice

Topic tags
Python
Data structures
Product design

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Dot product of two long integer vectors

Read a shared length and two integer vectors of that length, then return their dot product as a decimal string. Most entries are zero, so the useful work should follow the non-zero positions rather than the dec…

Practice focusRequirements - Line 1 is the length n ; line 2 holds n integers for the first vector; line 3 holds n integers for the second. - Some inputs carry the two characters backslash and n instead of real line breaks; treat both forms as line separators before parsing. - Multiply matching positions, then add the products. Entries may be negative, so the sum can be negative or zero. - Return the total as a plain decimal string with no padding, no plus sign, and no newline. - Do not print the individual products or the operand vectors.…

Previous4 of 11Next

Browse all original practice

Build a token-budget batcher
Sliding-window rate limiter
Interval calendar
Nested dictionary transforms
Command-log store
Dependency-ready scheduler
Streaming metric window
Evaluation slice aggregator
Build a reliable job runner
Build one adaptable decision story
Defend your role fit
Prepare one project deep dive
Design an evaluation artifact registry
Defend a reversible launch decision
Negotiate an ambiguous role scope
Trace a project from claim to evidence
Choose between two urgent priorities
Resolve a cross-functional disagreement
Own a consequential failure
Increase a team’s leverage
Warm a cold cache without a thundering herd
Ship a 500 GB checkpoint without a startup stampede
Make every document decision traceable to evidence
Share eight GPUs between two workloads with different promises
A mutable indexed sequence that must restart exactly
Trace a falling unit price to a rising bill
Keep a very long conversation usable while it streams
Design durable scheduled publishing
Make speculative code edits safe to accept
Recover a grouping rule from two-word phrases

CONTENT BOUNDARY

Independent frontier AI interview practice. Includes adaptations of collected interview material and original exercises; not official company questions and not affiliated with the companies represented. Collected material does not predict future interview questions. Report a correction.

Loading the question catalog…

FREE PRACTICE

Public demos, always pinned.

Start with either demo. They remain here when you filter or sort.

Free account: sign in to practice the marked Free questions.

Newest published questions; deterministic, never randomized.

Demo 1Number search with delayed feedback
Demo 2Design real-time device monitoring and control

Question bank filters

Reset filters

Question groups

Filter question bank by interview groupAll 269 Coding 162 System design 48 Behavioral 29 Hiring manager 16 Project deep dive 14

Topics

Independent and not affiliated with the companies represented. Collected material does not predict future interview questions.

245 questions

Sort question bankSortFree first Newest Oldest AI ratio ↓AI ratio ↑

Question bank results

Every question has a public brief with the task, constraints, and follow-ups. Labels describe interactive practice access.

Free practice

Topic tags
Communication

xAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Project walkthrough and motivation

Prepare one project story and one reason for the role. The project story is five to eight minutes. Open on a constraint and the choice you made.

Free practice

Topic tags
Storage & state
Communication

AnthropicBehavioralHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Decision under uncertainty

Free practice

Topic tags
ML / AI
Communication

AnthropicHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Safety and speed trade-off

Free practice

Topic tags
Reliability
Communication

AnthropicProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Reliability and recovery project

Walk through a system you made more reliable after a serious failure. Cover the failure boundary, your diagnosis, the design change you owned, the rollout, and the evidence that recovery improved.

Free practice

Topic tags
Storage & state
Product design
Communication

OpenAIBehavioralHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Cross-functional disagreement

Free practice

Topic tags
Storage & state
Reliability
ML / AI
Product design
Communication

OpenAIHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Product judgment under ambiguity

Free practice

Topic tags
Storage & state
Communication

OpenAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Cost and latency project

Walk through a project where you reduced latency or cost without weakening correctness. Cover the baseline, bottleneck, alternatives, the change you owned, rollout safeguards, and measured result.

Free practice

Topic tags
Reliability
Product design

OpenAISystem designHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Tool execution platform

Free practice

Topic tags
Storage & state
Security
Communication

xAIHiring managerHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

High-ownership role fit

Free practice

Topic tags
Storage & state
ML / AI
Communication

xAIProject deep diveHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Scaling bottleneck project

Walk through a project where scale exposed a bottleneck. Cover the workload, the measured limit, your model of the bottleneck, the intervention you owned, and the evidence that the limit moved.

Free practice

Topic tags
Data structures
ML / AI
Communication

xAISystem designHalf AI-generated

Discussion / designResponse prompt

Reported interview periodNot available

Real-time inference gateway

Free practice

Topic tags
Python

AnthropicCodingHalf AI-generated

Python codingWebAssembly + public tests

Reported interview periodNot available

Build an Unbiased Randomness Adapter

Turn an unknown biased bit source into balanced choices, a sampled load estimate, and isolated concurrent batches.

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

First position of a target in a sorted array

Given a nondecreasing array and a target value, report where the target first appears, or -1 when it is missing. A logarithmic search is expected, even though the same value may occur many times.

Free practice

Topic tags
Python
Concurrency

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Per-function exclusive runtime from a call log

A single thread records start and end events for nested function calls. For each function id, work out how many time units it actually spent running, excluding the units spent inside calls it made.

Free practice

Topic tags
Python
Data structures

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Maximum haul from non-adjacent houses

Each house on a street holds a known amount of money, and robbing two neighboring houses on the same night trips the alarm. Choose houses with no two adjacent to collect the largest possible total.

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Values unique to the first list

Two Python-style lists arrive as literal assignments. Collect every value that belongs to the first list but not the second, keeping the first list's order.

Free practice

Topic tags
Python
Algorithms

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Longest run of distinct characters

Given a string, measure the longest contiguous stretch in which no character repeats. Report only the length of that stretch.

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Peak simultaneous bookings

Given a list of half-open time ranges, report the largest number of ranges that are active at the same instant.

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Merge three ordered lists without repeats

Combine three nondecreasing integer lists into one nondecreasing list in which every distinct value appears once, then print it as a single space-separated line.

Free practice

Topic tags
Python
Data structures

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Next arrangement in lexicographic order

Rearrange an integer list into the next larger lexicographic arrangement, wrapping around to the smallest arrangement when no larger one exists.

Free practice

Topic tags
Python

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Count distinct values in an ordered list

Given a nondecreasing integer sequence, report how many different values it contains.

Free practice

Topic tags
Python
Algorithms

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Zero rows and columns around each zero

Given a matrix, set every cell in a row or column that contains a zero to zero, deciding which rows and columns to clear from the original values rather than from the partially rewritten matrix.

Free practice

Topic tags
Python
Storage & state

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Normalize an absolute filesystem path

Read one absolute path and return its canonical form: empty segments and . segments disappear, .. folds into the parent, and exactly one slash separates the surviving directories.

Free practice

Topic tags
Python
Data structures
Product design

MetaCoding

Python codingWebAssembly + public tests

Reported interview periodNot available

Dot product of two long integer vectors

Previous4 of 11Next

Browse all original practice

Build a token-budget batcher
Sliding-window rate limiter
Interval calendar
Nested dictionary transforms
Command-log store
Dependency-ready scheduler
Streaming metric window
Evaluation slice aggregator
Build a reliable job runner
Build one adaptable decision story
Defend your role fit
Prepare one project deep dive
Design an evaluation artifact registry
Defend a reversible launch decision
Negotiate an ambiguous role scope
Trace a project from claim to evidence
Choose between two urgent priorities
Resolve a cross-functional disagreement
Own a consequential failure
Increase a team’s leverage
Warm a cold cache without a thundering herd
Ship a 500 GB checkpoint without a startup stampede
Make every document decision traceable to evidence
Share eight GPUs between two workloads with different promises
A mutable indexed sequence that must restart exactly
Trace a falling unit price to a rising bill
Keep a very long conversation usable while it streams
Design durable scheduled publishing
Make speculative code edits safe to accept
Recover a grouping rule from two-word phrases

CONTENT BOUNDARY