Anthropic OA Questions 2025: What Candidates Actually Saw

Last updated: May 2026

8+ Anthropic online assessment questions reported by candidates who took the OA in 2025. Sourced from 1Point3Acres, LeetCode, Blind, Glassdoor, and Reddit. Data grows as new reports are added.

Anthropic OA Format in 2025

The Anthropic online assessment has followed a consistent format across multiple hiring cycles. Based on 2025 reports in the database, candidates typically see 2 to 3 algorithmic coding problems delivered through a proctored online platform (most commonly HackerRank). Some roles include additional behavioral or work-simulation components after the coding section.

Time limits reported by candidates range from 70 to 110 minutes for the full assessment. The coding problems are auto-graded, so edge case handling and runtime efficiency both matter. Partial credit is available on some platforms if your solution passes a subset of test cases.

Real Anthropic OA Questions from 2025

Anthropic Online Test Interview Experience for SDE Position

oa,ml,distributed-systems,mode-finding,data-batcher,rl,grpo,debugging 1p3a

This post was last edited by Jing666 on 2025-10-3 15:25. Looks like I failed. Coding: Q4, interview question: distributed worker find mode, but I didn't have time to write the follow-up find median, o

View

Anthropic SDE Online Test Experience and C++ Interview Language Query

hard oa,database,sql,time-travel-query,c++,codesignal 1p3a

This post was last edited by Anonymous on 2025-10-05 18:44. The questions are below. Has anyone had an interview using C++? The HR called and told me I had to choose a language beforehand. I'm more fa

View

Cloud Storage System (Online Assessment)

easy oa,design,simulation,sorting,strings,prefix-suffix,user-management 1p3a

Problem Summary Your goal is to build a simple, in-memory cloud storage system. This system will store files and their details, such as name and size. **Important:** This system runs entirely in me

View

Employee Management System (Online Assessment)

easy oa,design,simulation,heap,sorting,payroll 1p3a

Problem Summary You need to build a simple system to manage employees. This challenge has **4 parts**. Each part gets a little harder than the last: 1. **Part 1**: Add new employees and track when

View

In-memory Database (Online Assessment)

easy oa,database,simulation,ttl,time-travel-query,compare-and-swap,strings 1p3a

Problem Overview You need to build a simple in-memory database. This challenge has **4 levels**. You must solve the current level to unlock the next one. 1. **Level 1**: Basic tools to add, update,

View

Task Management System (Online Assessment)

easy oa,design,simulation,sorting,strings,data-structures 1p3a

Task Management System (Interview Problem) ## Problem Summary You need to build a task management system. There are **4 levels**. Each level gets harder and adds new features: 1. **Level 1**: Build

View

Banking System (Online Assessment)

easy oa,design,simulation,sorting,data-structures,heap 1p3a

System Overview You need to build a bank system that runs in the computer's memory. There are **4 levels** to this problem. You must finish one level to move to the next. 1. **Level 1:** create acco

View

Recipe Manager (Online Assessment)

easy oa,design,simulation,sorting,hash-table,versioning,user-management 1p3a

Problem Summary You need to build a recipe management system that runs in memory. The problem has **4 levels**. Each level adds new features to the previous one: 1. **Level 1**: Basic actions (Crea

View

Topic Breakdown (2025 OA)

Topics reported by candidates who took the Anthropic OA in 2025:

oa,ml,distributed-systems,mode-finding,data-batcher,rl,grpo,debugging (1) oa,design,simulation,sorting,strings,prefix-suffix,user-management (1) oa,design,simulation,sorting,strings,data-structures (1) oa,design,simulation,sorting,hash-table,versioning,user-management (1) oa,design,simulation,sorting,data-structures,heap (1) oa,design,simulation,heap,sorting,payroll (1) oa,database,sql,time-travel-query,c++,codesignal (1) oa,database,simulation,ttl,time-travel-query,compare-and-swap,strings (1)

Difficulty Breakdown (2025 OA)

6

easy

1

hard

Difficulty labels are reported by candidates or inferred from LeetCode ratings.

How to Prepare for the Anthropic OA

The Anthropic OA is a timed, auto-graded coding test. Preparation is different from interview prep: you need to optimize for both correctness and speed, since partial credit requires passing test cases, not just showing reasoning.

  • Practice timed sessions. Set a timer for 35 to 45 minutes per problem and simulate the real constraint.
  • Focus on the topic clusters that appear most in Anthropic OA reports: dynamic programming, graphs, arrays, and strings make up the majority of reported problems.
  • Read the problem constraints carefully before writing code. The constraints often reveal the expected time complexity and algorithm type.
  • Use the LeakCode database to review 2025 OA questions from actual candidates. Pattern recognition across multiple sightings is a stronger prep signal than any single mock problem.
  • If the OA includes a work simulation section, practice under realistic conditions. This section tests judgment and values alignment, not coding ability.

FAQ

What does the Anthropic OA look like in 2025?

Based on 2025 candidate reports, the Anthropic OA typically includes 2 to 3 coding problems and runs 70 to 110 minutes. Topics lean toward dynamic programming, graphs, arrays, and strings. Some roles include a work simulation section after the coding component.

How long is the Anthropic OA?

Most candidates report 70 to 90 minutes for the coding component. Roles with additional work simulation sections extend the total time to 110 to 150 minutes. Check your invitation email for the exact time allocation for your specific role.

What platform does Anthropic use for the OA?

Anthropic uses HackerRank or its own proprietary platform depending on the role and hiring cycle. Your invitation will specify the platform. Both platforms support multiple languages including Python, Java, C++, and JavaScript.

Are these real Anthropic OA questions?

Yes. All questions in the LeakCode database come from candidate reports on 1Point3Acres, Blind, Glassdoor, Reddit, and LeetCode. They are not AI-generated. Each question links to its source where available.

Other Anthropic OA Years

Related: Anthropic OA Hub · Anthropic All Questions · All OA Questions · Data Sources