CCAR-F Prompt Engineering and Structured Output Practice Question
You notice that your prompt, which works perfectly with Claude 3.5 Sonnet, fails when you switch to Claude 3 Haiku. What is the most likely reason?
⚠ Common exam trap
Candidates often attribute the failure to a bug in the API or the model itself, rather than recognizing the reduced instruction-following capability of smaller, faster models.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Haiku requires more granular instructions and more few-shot examples.
Smaller models like Haiku have less 'instruction-following capacity' and may require more explicit, simplified prompting compared to larger models like Sonnet. When switching to smaller models, you must often reduce task complexity, provide more examples, and simplify formatting instructions. The smaller model's reduced parameter count necessitates clearer, less ambiguous prompts to achieve the same level of structural adherence and task accuracy expected from larger, more capable models.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
The context window is smaller on Haiku.
Why it's wrong here
While model capacities differ, both Sonnet and Haiku offer substantial context windows. The failure is rarely due to raw window size but rather the model's ability to process the instructions embedded within that window. The issue lies in the complexity of the instructions being too dense for the smaller model.
- ✗
Haiku has a higher temperature default, causing more errors.
Why it's wrong here
Temperature settings are controlled by the user, not the model version itself. While different models have different baseline sensitivities, changing the model does not implicitly change the temperature. The performance gap relates to the model's inherent reasoning and instruction-following capabilities, not the sampling parameters configured by the user.
- ✓
Haiku requires more granular instructions and more few-shot examples.
Why this is correct
Smaller models benefit significantly from explicit, step-by-step instructions and robust few-shot examples to compensate for lower parameter counts. What suffices as a 'hint' for Sonnet may be insufficient for Haiku. Refining the prompt to be more prescriptive will help the smaller model perform the task more effectively.
- ✗
The API endpoint has changed for the smaller model.
Why it's wrong here
Anthropic's API endpoints remain consistent across different model versions. The issue is not the connectivity or endpoint structure, but the interaction between the prompt's linguistic complexity and the model's internal capability. This is an architectural prompt engineering challenge rather than a technical API integration or infrastructure failure issue.
About these practice questions
This CCAR-F question is part of Courseiva's 271-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Anthropic exam blueprint
This CCAR-F practice question is part of Courseiva's free Anthropic certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the CCAR-F exam.