The AI lab interview round that asks the most questions isn't in the onsite.

Land the Lab ·

34% of reported questions at the top AI labs come from the technical screen, more than from any onsite round. 9 loops compared: rounds, question types and AI-tool rules.

A person working through formulas on a chalkboard
Photo: Vitaly Gariev on Unsplash
  • 34% of reported questions come from the technical screen, more than any onsite round.
  • 4 to 8 rounds; 7 is typical. Cursor, part of SpaceXAI (formerly xAI), runs the shortest loop (4) around a one- or two-day onsite project.
  • 54% of questions are coding. Mistral AI and Cohere are the exceptions: ML design leads there.
  • AI tools split the labs. Not allowed at OpenAI, Google DeepMind, Perplexity and Cursor; allowed at Cohere.

Seven rounds is typical; Cursor runs 4

Rounds from the first recruiter call to the last interview, as candidates report them. Cursor is part of SpaceXAI (formerly xAI) but hires through its own loop, so it gets its own row.

  1. Anthropic8 rounds
  2. Scale AI8 rounds
  3. Cohere7 rounds
  4. Google DeepMind7 rounds
  5. OpenAI7 rounds
  6. Perplexity7 rounds
  7. SpaceXAI (xAI)7 rounds
  8. Mistral AI5 rounds
  9. Cursor (SpaceXAI)4 rounds

4 of 9 loops make you build something real

Every loop opens with a recruiter call and all but one have a technical screen. Four have a practical coding round, where you build or fix something real: OpenAI, SpaceXAI (xAI), Scale AI and Cursor (SpaceXAI).

CompanyOnline test / take-homeTechnical screenCodingPractical codingSystem designAI / ML designProject deep diveValues / behavioralHiring manager
OpenAInoyesnoyes2noyesyesno
Anthropicyesyesyesnoyesnoyesyesyes
Google DeepMindnoyesyesnoyesnoyesyesyes
SpaceXAI (xAI)yesyesnoyesyesnoyesnoyes
Perplexitynoyesyesnoyesyesyesnoyes
Scale AIyesnoyes2noyesnoyesyes
Coherenoyesyesnonoyesnoyes2
Mistral AInoyesyesnoyesnonoyesno
Cursor (SpaceXAI)noyesnoyesnonoyesnono

More questions come up in the screen than in any onsite round

Share of 289 reported questions by the round they were asked in. The screen is where most loops end, so it is where the most reports come from.

  1. Technical screen34%
  2. Coding26%
  3. AI / ML design12%
  4. System design10%
  5. Online test / take-home6%
  6. Recruiter3%
  7. Practical coding3%

Coding leads everywhere except Mistral AI and Cohere

Question types per company. Google DeepMind is left out: too few reports to chart.

  • Coding
  • ML design
  • System design
  • AI system design
  • Behavioral
  1. OpenAI n=116Coding 53%ML design 6%System design 23%AI system design 9%Behavioral 8%
  2. Anthropic n=67Coding 48%ML design 13%System design 12%AI system design 16%Behavioral 10%
  3. SpaceXAI (xAI) n=51Coding 71%ML design 6%System design 8%AI system design 8%Behavioral 8%
  4. Perplexity n=25Coding 64%ML design 4%System design 16%AI system design 12%Behavioral 4%
  5. Scale AI n=25Coding 68%ML design 16%System design 8%Behavioral 8%
  6. Mistral AI n=16Coding 19%ML design 50%AI system design 25%Behavioral 6%
  7. Cohere n=13Coding 31%ML design 38%System design 15%AI system design 15%
  8. Cursor (SpaceXAI) n=13Coding 46%System design 23%AI system design 15%Behavioral 15%

AI tools: not allowed at 4 companies, allowed at 1

Not allowed

  • OpenAINo AI assistants or autocomplete in the coding screen or the practical round
  • Google DeepMindUnaided; AI not allowed or heavily restricted
  • PerplexityAssume no AI unless the recruiter says otherwise
  • Cursor (SpaceXAI)No AI other than autocomplete in the first screens, its CEO says; the onsite project's rules are not confirmed

Depends on the round

  • AnthropicNo AI in the online test, screens, coding or system design; some 2026 loops add a round coded with Claude Code
  • SpaceXAI (xAI)AI expected in the take-home; none in the phone screen
  • Mistral AIReportedly fine for boilerplate and lookups; fundamentals must be yours

Allowed

  • CohereGenerally allowed; graded on the work, not on how you use AI

Not stated

What this means for your prep

  1. Treat the screen as the interview. More reported questions are asked there than in any other round; practice multi-part problems that grow every 15 minutes.
  2. Practice building, not just solving. At OpenAI, SpaceXAI (xAI), Scale AI and Cursor (SpaceXAI), a round is about working code you would ship.
  3. Match the mix. Interviewing at Mistral AI or Cohere? Spend more time on ML design than on algorithms.
  4. Ask about AI tools in the recruiter call. Practice both ways if the answer is unclear.

See every reported question for your loop, by round, with when it was last asked.

Pick your company