CCAO-F - Product and Model Selection (12% of the exam) - Section 3.3

Align model selection with task requirements (cost, speed, quality).

Choosing a model for the job at hand: a faster, lower-cost model for high-volume, straightforward work and a more capable model for complex reasoning. The guide's sample item turns on exactly this for high-volume reply drafts.

high-volume simple taskscomplex reasoning taskscost, speed and quality trade-off

Practice question for this objective

Free sampleProduct and Model Selectionmedium

An operations manager at a logistics company finds her team uses the most capable model for everything, including tidying one-line meeting notes and drafting routine shift-swap messages. During busy morning handovers these replies are slow, and the company's AI guidance asks staff to match the model to the task. The team also uses Claude for occasional complex route-planning analysis. What should she change?

AKeep the most capable model for every task, because better quality outweighs a short wait during handovers
BMove the whole team to the fastest model for everything, including the complex route-planning analysis
CSteer routine notes and messages to a faster, lower-cost model and keep the top model for complex planning Correct
DAsk the team to write longer, more detailed prompts so the most capable model returns its answers faster

Always defaulting to the most capable model wastes time and cost on simple work; route routine tasks to faster models and reserve the top model for complex ones. Each model tier trades capability against speed and cost. Simple, frequent tasks gain nothing from the most capable model but still pay its slower responses and higher cost, which compound across a busy team. Matching tiers to tasks keeps routine work quick and cheap while preserving deeper reasoning for the analysis that needs it.

Why A is wrong: Defaulting to the top model feels like the safe choice. For one-line tidy-ups it adds delay and cost with no visible quality gain, and it ignores the company's guidance to match the model to the task.

Why B is wrong: This fixes the slow mornings and cuts cost. It over-corrects: the route-planning work genuinely needs deeper reasoning, so pushing it onto the fastest model risks weaker analysis.

Why C is correct: This matches each kind of work to the model it needs: quick, simple items go to a faster, cheaper model, while the occasional complex analysis keeps the most capable one. It follows the company guidance and removes the morning delay.

Why D is wrong: Better prompts often improve results, so this sounds sensible. Longer prompts do not make a model respond faster, and the real issue is using a slow, costly model for trivial work.

See more CCAO-F practice questions, answers explained.

Exam traps in Product and Model Selection

Answers that look right on this material and are not. Each one is a distractor from a different question in the CCAO-F bank for this domain.

Use the most capable model for both tasks so the quality of every output is as high as it can be
Why it is wrong: Defaulting to the top model feels like a quality guarantee. For 200 short post variants it adds cost and wait time without a meaningful gain, which is exactly what the stem says matters for that task.
Use the most capable model for every description, because the programme will be seen by every delegate
Why it is wrong: A public audience makes quality feel paramount, so the top model is tempting. But a simple reformat needs no deep reasoning, and the most capable model is slower and costlier across 400 items with a Friday deadline.
Stay with the fastest model and regenerate each segment until the tone finally sounds right
Why it is wrong: Regenerating is quick and costs little each time. Re-running an unchanged, already well-structured prompt on a model that falls short for the task is hoping rather than fixing, and it eats the time she is trying to save.

Examworthy is not affiliated with or endorsed by Anthropic. Original, blueprint-aligned practice material only.