How Anthropic Interviews in 2026: Process, Questions & What They Score

1. 01

   Build out core business logic for a banking system that must handle multiple transaction types, with requirements added incrementally as the interview progresses.

   Tests incremental implementation: core logic that absorbs new transaction-type requirements without rewrites.

2. 02

   Write a function that converts a list of profiler stack samples (timestamps + call stacks) into a list of start/end trace events for each function call.

   Tests parsing stack samples into start/end events; the insight is diffing adjacent stacks for entries and exits.

3. 03

   Given stack trace samples, determine the longest-running function, including handling edge cases like recursive or infinite-loop-like patterns.

   Tests interval accounting over stack samples, with edge cases like recursion and never-returning frames.

4. 04

   Implement a data structure or algorithm from scratch (e.g., a specific hash map or parser) and be ready to defend your time/space complexity choices under questioning.

   Tests from-scratch implementation of a hash map or parser and defending time/space choices under questioning.

5. 05

   (Research Engineer variant) Implement a Transformer component such as multi-head attention from scratch in PyTorch.

   Tests PyTorch fluency: tensor shapes, projections, scaled dot-product attention, and head splitting from scratch.

6. 06

   (Research Engineer variant) Debug a broken/incorrect model training loop and explain what's wrong and why.

   Tests ML debugging: spotting issues like missing zero_grad, wrong loss reduction, or data leakage, and explaining why.

7. 07

   Refactor your existing solution cleanly as new constraints or requirements are layered on mid-interview, without a full rewrite.

   Tests refactoring discipline: absorbing new constraints cleanly mid-interview without a full rewrite.

Reported problems, listed so you know what to expect. Practice them in your own editor or on the platform you prefer; SupaCV's practice mode coaches how you talk through them.

1. 01

   Design an inference batching system for a single GPU that can process up to 100 inputs per batch, where users submit requests synchronously and wait for results.

   System design: batching policy and the latency versus throughput tradeoff with synchronous callers waiting.

2. 02

   Design an end-to-end request pipeline for LLM queries: intake, batching, GPU routing/load balancing, and returning responses to the correct caller.

   System design: request intake, batching, GPU routing and load balancing, and returning responses to the right caller.

3. 03

   Design a token-generation service that must handle very high request volume: reason about throughput, horizontal scaling, and request routing.

   System design: throughput estimation, horizontal scaling, and request routing under very high load.

4. 04

   How would you manage GPU memory (e.g., KV cache) under many concurrent, variable-length generation requests?

   Tests GPU memory management: KV-cache sizing, paging or eviction, and admission control for variable-length requests.

5. 05

   How would you support streaming responses and a priority-based request queue in an LLM-serving system?

   System design: streamed token delivery plus priority scheduling and fairness across request classes.

6. 06

   Design the backend architecture for a consumer-facing chat product like Claude.ai.

   System design: chat persistence, streaming responses, model-serving integration, and scaling stateless layers.

Reported problems, listed so you know what to expect. Practice them in your own editor or on the platform you prefer; SupaCV's practice mode coaches how you talk through them.

1. 01

   Why do you want to work at Anthropic specifically, as opposed to another AI lab (e.g., OpenAI, Google DeepMind)?

2. 02

   What are your thoughts on AI safety and the risks posed by advanced AI systems?

3. 03

   Which of Anthropic's stated values resonates most with you, and why?

4. 04

   What are the ethical risks of deploying agentic AI systems in high-stakes environments?

5. 05

   How would you handle being assigned to work on a project you personally believed was unsafe or premature to ship?

6. 06

   What's a specific, considered (not generic) point of view you hold about how AI development should be governed or paced?

1. 01

   Tell me about a time you did something that conflicted with your own values.

2. 02

   Tell me about a time you pushed back on a decision and lost: how did you handle it afterward?

3. 03

   Describe a time you changed your mind about something you previously felt strongly about.

4. 04

   Tell me about a time you worked on something you had moral or ethical reservations about.

5. 05

   Walk me through a project you're proud of and the tradeoffs you navigated to get there.

6. 06

   Who do you respect but disagree with on values, and why?

7. 07

   Describe a time you made a decision under significant uncertainty with no clearly right answer.