Answers for all three versions
Every version has the same 45 questions and the same answer choices. Only the order of the questions, and the order of the choices in each question, is different.
Find your version by question 1 of your test paper:
- Version A: question 1 starts “The task app is written with Next.js, which is a framework. What is a framework?”
- Version B: question 1 starts “What does it mean to distill an AI model?”
- Version C: question 1 starts “You have step-by-step instructions, plus a small program, for turning your …”
Each row is one question. The link goes to its explanation (written for Version A: the letters of the choices may be different in your version, but the text of the choices is the same).
| Version A | Version B | Version C | Question | Explanation |
|---|---|---|---|---|
| 1: a | 42: e | 28: c | The task app is written with Next.js, which is a framework. What is … | Question 1 |
| 2: c | 32: b | 22: c | A startup is choosing a programming language for a new product. What … | Question 2 |
| 3: b | 33: d | 24: a | Which part of the task app is its front end? | Question 3 |
| 4: a | 5: d | 34: d | What is recursive self-improvement (RSI)? | Question 4 |
| 5: c | 40: c | 23: a | How do strong open-weight models affect the business of Anthropic … | Question 5 |
| 6: a | 7: c | 40: b | Why is a prompt cache useful? | Question 6 |
| 7: d | 29: d | 12: c | Which choice correctly describes both an LLM and an agent? | Question 7 |
| 8: c | 26: d | 7: d | Which is an example of agentic AI, not just generative AI? | Question 8 |
| 9: d | 35: c | 32: c | A student on a paid Claude plan is suddenly blocked from using … | Question 9 |
| 10: d | 18: b | 27: a | Which choice correctly describes both AGI and superintelligence? | Question 10 |
| 11: d | 39: c | 5: c | Why have AI companies such as Anthropic and Google worked so hard to … | Question 11 |
| 12: c | 23: b | 6: a | Which of these takes up space in Claude Code’s context window, even … | Question 12 |
| 13: d | 34: d | 13: b | Which of the following are in Claude Code’s context window for a … | Question 13 |
| 14: b | 24: a | 41: e | After a very long Claude Code session, Claude starts ignoring a … | Question 14 |
| 15: c | 27: d | 33: d | Why do LLMs work so well with Markdown files, compared with Word or … | Question 15 |
| 16: d | 22: b | 14: c | During training, a model reads a sentence and predicts the next … | Question 16 |
| 17: d | 25: a | 15: b | Claude Code uses the Claude Opus 5.5 model. Which describes their … | Question 17 |
| 18: b | 1: a | 17: d | What does it mean to distill an AI model? | Question 18 |
| 19: a | 9: d | 3: c | You need the exact sum of two 100-digit numbers. You try two ways: … | Question 19 |
| 20: b | 19: c | 26: b | Claude Code reads a web page while it helps you. The page has white … | Question 20 |
| 21: e | 2: b | 1: c | You have step-by-step instructions, plus a small program, for … | Question 21 |
| 22: c | 12: a | 29: b | Our static website’s repo was already on GitHub. Then we turned on … | Question 22 |
| 23: a | 37: d | 38: d | Why is it risky to give a very powerful AI a goal such as “Make as … | Question 23 |
| 24: a | 11: a | 36: a | You send Claude a question that is 1,000 tokens long, and it replies … | Question 24 |
| 25: a | 38: b | 37: b | A hospital wants AI to summarize patient records, but it must not … | Question 25 |
| 26: d | 43: e | 20: e | A tokenizer splits the text “Elephants are enormous.” into tokens. … | Question 26 |
| 27: d | 8: b | 8: a | A browser asks a web server for a page. To build the page, the web … | Question 27 |
| 28: a | 16: a | 21: b | What is the difference between /compact and /clear in Claude Code? | Question 28 |
| 29: e | 36: c | 31: d | A bank wants its AI to stick closely to the most probable words, so … | Question 29 |
| 30: c | 41: d | 9: b | You want Claude Code to follow a rule every time it works on your … | Question 30 |
| 31: b | 13: a | 43: a | Which of these could be built as a static website (only HTML, CSS … | Question 31 |
| 32: d | 14: a | 4: b | Which task is the best fit for an agentic AI tool, such as Muse, and … | Question 32 |
| 33: b | 3: d | 11: d | A Mac user, a Linux user and a Windows user all want to use the Bash … | Question 33 |
| 34: a | 4: c | 42: d | Three students work together on the same project. What role does … | Question 34 |
| 35: e | 30: c | 19: e | What is Vercel’s job in the task app’s architecture? | Question 35 |
| 36: e | 15: c | 35: c | Model X has a few billion parameters. Model Y has over a trillion. … | Question 36 |
| 37: b | 20: a | 30: c | A web page has (1) headings, paragraphs and a bullet list; (2) its … | Question 37 |
| 38: b | 17: b | 16: a | An LLM is completing “The boy went to the ___“. Suppose”store” is … | Question 38 |
| 39: c | 6: e | 39: e | What is the /rewind command in Claude Code used for? | Question 39 |
| 40: c | 10: b | 2: c | What is the significance of a model’s training cutoff date? | Question 40 |
| 41: b | 21: e | 18: d | In Claude Code you have two jobs: renaming 50 files, and redesigning … | Question 41 |
| 42: a | 28: c | 25: b | Your project’s code worked last week, but today it is broken. How … | Question 42 |
| 43: c | 31: c | 10: a | You move a file to another folder by typing a command in a terminal. … | Question 43 |
| 44: 1A, 2G, 3C, 4E, 5B | 45: 1A, 2C, 3G, 4B, 5D | 45: 1F, 2D, 3B, 4G, 5E | Match each term with its meaning. Two meanings are not used. (10 … | Question 44 |
| 45: 1B, 2C, 3A, 4E, 5F | 44: 1F, 2C, 3A, 4D, 5B | 44: 1B, 2F, 3D, 4E, 5C | Match each part of the task app’s tech stack with its job. One job … | Question 45 |
Answer key by version
The same answers, listed by question number in each version.