CCAO-F Claude Model Fundamentals — 30 Questions

Question 1

When designing a system that requires Claude to output structured data (JSON), which approach is most reliable for ensuring the output is valid and parsable?

Accepted Answer

Include a system prompt instruction to output valid JSON and pre-fill the assistant response with '{'.. Using a combination of a clear schema definition in the system prompt and a 'pre-fill' strategy—where the assistant message starts with a brace—is the industry standard. This technique constrains the model's output generation to match the expected format. Mastering structured output is critical for building robust pipelines where Claude's output must be immediately consumed by downstream software components without manual cleaning or parsing failures.

Answer

Ask the model to generate the JSON and assume it will be valid every time.

Answer

Wrap the response in XML tags and ask the user to manually copy it.

Answer

Set the temperature to 0.0 and hope for the best.

Question 2

Which TWO of the following are recommended practices for managing Claude's token usage in long-context applications?

Accepted Answer

Summarize old conversation turns that are no longer strictly necessary.. Efficient token management is essential for controlling API costs and maintaining low latency. By proactively summarizing conversation history and selectively including only relevant context from retrieved documents, developers can keep the total token count within optimal ranges. These practices ensure the application remains cost-effective and performant, which is a core requirement for scaling AI-driven features in high-traffic enterprise production environments.

Answer

Send the entire conversation history in every single API call.

Answer

Use the largest possible max_tokens setting for every request to avoid truncation.

Answer

Avoid using XML tags to keep the token count as low as possible.

Question 3

Which THREE factors significantly influence the latency of a response from the Claude API?

Accepted Answer

The length and complexity of the input prompt.. Latency is influenced by both the complexity of the task and the amount of data processed. Larger context inputs require more computation, and longer output tokens directly increase generation time. Furthermore, network conditions impact the delivery of these tokens. Optimizing these factors is necessary for building responsive, real-time AI applications that satisfy user expectations for speed and performance in modern web and mobile environments.

Answer

The temperature setting of the request.

Answer

The number of previous API calls made by the account.

Question 4

Which of the following is the recommended approach for handling long-running conversational history in the Anthropic Messages API?

Accepted Answer

Store the history in a database and summarize it before each request.. To maintain optimal performance and cost-efficiency, developers should implement a sliding window or summarization strategy for message history. Since the API requires sending the entire conversation history with each request, keeping the history concise prevents hitting token limits and minimizes the overhead per API call. This architectural approach ensures that applications remain responsive and cost-effective as conversations grow significantly longer over time in production environments.

Answer

Send the entire conversation history in every single API call.

Answer

Use an external vector database for all past messages.

Answer

Rely on the model's inherent memory to track the chat.

Question 5

When implementing a 'Human-in-the-Loop' workflow with Claude, what is the best practice for handling user feedback on generated content?

Accepted Answer

Send feedback to the model in the next conversation turn.. The most effective approach is to collect user feedback as structured data and incorporate it into the next iteration of the prompt or conversation history. This allows the model to adjust its behavior based on specific user requirements or factual corrections. Explicitly feeding this feedback back into the context cycle turns a passive interaction into an active learning loop, directly improving the quality and relevance of the model's future responses.

Answer

Ignore user feedback to prevent model bias.

Answer

Automatically generate a new prompt from feedback.

Answer

Use the feedback to retrain the base model.

Question 6

A developer needs to provide Claude with a 150,000-word technical manual to help answer support questions. What is a fundamental limitation of the Claude 3 model family's context window that the developer must consider?

Accepted Answer

Tokens are approximately 0.75 words, so 150,000 words may exceed 200,000 tokens.. While Claude 3 models have a large context window (200k tokens), a token is not equivalent to a word. Understanding the relationship between tokens and words is essential for capacity planning and ensuring that very large documents do not exceed the model's architectural limits during processing.

Answer

Claude 3 models have a context window limited to 32,000 tokens.

Answer

The context window only applies to the model's output, not the input.

Answer

Vision inputs do not count toward the token limit in the context window.

Question 7

When designing an application using Claude 3.5 Sonnet, which TWO advantages does this model offer over the original Claude 3 Sonnet? (Select TWO)

Accepted Answer

Significant improvements in coding and nuance. Claude 3.5 Sonnet represents a mid-generation update that significantly improves performance metrics without increasing the model's size or cost. Identifying these specific improvements is crucial for developers deciding whether to migrate existing workflows from the 3.0 version to the 3.5 version.

Answer

A larger context window of 1 million tokens

Answer

Introduction of vision capabilities for the first time

Answer

Support for 100+ additional spoken languages

Question 8

A developer is building a coding assistant. Which TWO features of Claude 3.5 Sonnet make it particularly well-suited for this use case compared to previous generations? (Select TWO)

Accepted Answer

State-of-the-art performance on coding benchmarks like HumanEval. Claude 3.5 Sonnet introduced significant advancements in technical reasoning. These improvements are not just general intelligence boosts but specific architectural refinements that allow the model to better understand complex codebases and follow multi-step instructions, which are essential for developer tools and automated programming tasks.

Answer

Ability to execute Python code in a local sandbox natively

Answer

Unlimited free usage for all API developers

Answer

Built-in integration with all major IDEs natively

Question 9

Which TWO statements accurately describe how Claude handles system instructions compared to user messages? (Select TWO)

Accepted Answer

System instructions are provided in a separate 'system' parameter.. The distinction between system instructions and user messages is a core concept in the Messages API. System instructions provide the 'global' context for the model's behavior, while user messages provide the 'local' context for the specific task at hand. Understanding their interaction is key to effective prompt engineering.

Answer

User messages always override system instructions if they conflict.

Answer

System instructions are mandatory for every API call.

Answer

System instructions must be written in XML to be understood.

Question 10

A software engineer is deciding between Claude 3 Opus and Claude 3.5 Sonnet for a complex automated refactoring tool. Which TWO statements accurately describe the comparative advantages of Claude 3.5 Sonnet over Claude 3 Opus?

Accepted Answer

Claude 3.5 Sonnet operates at roughly twice the speed of Claude 3 Opus.. Claude 3.5 Sonnet represents a generational leap over the original Claude 3 family, including the Opus model. Despite being in the 'Sonnet' tier, the 3.5 version outperforms the older 3.0 Opus on many industry benchmarks, particularly in coding and logical reasoning. Furthermore, it maintains the speed and cost-efficiency characteristic of the Sonnet tier, offering a superior performance-to-price ratio. (71 words)

Answer

Claude 3.5 Sonnet has a larger context window than Claude 3 Opus.

Answer

Claude 3.5 Sonnet is the only model that supports system prompts.

Answer

Claude 3.5 Sonnet is more expensive per token than Claude 3 Opus.

Question 11

Which THREE types of visual data can Claude 3 models process as part of their native vision capabilities?

Accepted Answer

JPEG and PNG image files. The Claude 3 family is multimodal, meaning it can accept and interpret various image formats alongside text. This allows users to include visual evidence in their prompts, such as screenshots of code, photographs for identification, or digitized documents. Supporting these formats is essential for modern workflows that rely on more than just plain text for communication and data analysis. (70 words)

Answer

Live video streams (RTSP/RTMP)

Answer

3D CAD models (.STL or .OBJ)

Question 12

An organization wants to ensure their application uses the most stable and up-to-date version of Claude 3.5 Sonnet. They notice the model string 'claude-3-5-sonnet-20240620'. What does the suffix '-20240620' represent in this model identifier?

Accepted Answer

The specific release version of the model architecture.. Anthropic uses date-based versioning to identify specific releases of their models. This allows developers to pin their applications to a specific version, ensuring that updates to the model's underlying weights or behavior do not unexpectedly break their prompts or workflows. Using a specific date suffix provides a balance between innovation and reliability. (68 words)

Answer

The date when the model's training data was last updated from the internet.

Answer

The expiration date after which the model will no longer be supported.

Answer

The maximum context window limit for that specific model instance.

Question 13

The provided JSON exhibit results in output that occasionally includes conversational filler before the JSON block. What change should be made to ensure strictly valid JSON output?

Accepted Answer

Use the API's 'assistant' role to pre-fill the start of the response with '{'.. When Claude produces conversational filler, it is often because the system prompt is too broad. By adding a 'pre-fill' strategy or explicitly instructing the model to start with a specific character, you constrain the generation path. For JSON extraction, pre-filling the response with '{' forces the model to complete the structure, eliminating unwanted preamble and ensuring the output is immediately parseable by downstream systems.

Answer

Decrease the temperature to 0.0 to make the model more deterministic.

Answer

Add 'Do not include any conversational text' to the system prompt.

Answer

Increase the max_tokens parameter to ensure the JSON does not get truncated.

Question 14

Which THREE factors significantly influence the latency of a Claude API call?

Accepted Answer

The number of tokens requested in the max_tokens parameter.. Latency is a multi-dimensional challenge in LLM orchestration. The number of tokens requested directly impacts the compute time required for generation. Input size influences the initial processing phase (prefill), while the underlying hardware load on Anthropic's infrastructure and network round-trip times also play roles. Understanding these variables is vital for engineers designing high-performance systems that require real-time responses and predictable user experience across varying request volumes.

Answer

The complexity of the temperature setting used for the request.

Answer

The specific font used for the text input by the user.

Question 15

An engineer is streaming Claude's output to a web client and notices that the first visible token sometimes appears after a noticeable delay on long prompts. They want to reduce perceived latency without changing the model or the prompt content. Which change is most likely to help?

Accepted Answer

Enable streaming in the request so tokens are delivered incrementally as Claude generates them.. Streaming changes when tokens are delivered, not how they are generated, so the client renders the first token as soon as it is produced. This is the standard way to improve perceived responsiveness for long prompts. The other options either affect output length, model cost, or sampling behavior, none of which shorten the wait before the first token.

Answer

Switch to a larger Claude model so the response is generated more accurately.

Answer

Reduce the temperature to 0 so Claude decides on the first token faster.

Answer

Increase the max_tokens value so Claude has more room to generate the full answer.

Question 16

A financial analyst uses Claude to extract data from scanned PDF invoices. The invoices sometimes contain handwritten notes in the margins. The analyst needs the extracted data to be accurate and needs to know when Claude is uncertain about a field. Which approach best meets these requirements?

Accepted Answer

Instruct Claude to return a confidence score (0-100) for each extracted field and flag any field with a score below a threshold for manual review.. Requesting per-field confidence scores allows the analyst to systematically identify uncertain extractions and route them for human review. This leverages Claude's ability to self-evaluate and produce structured output, ensuring both accuracy and transparency. Other options either lack a direct uncertainty measure or rely on unstructured notes that are harder to act upon programmatically.

Answer

Use Claude to extract the data and then run a separate OCR tool on the same invoices to cross-validate the results.

Answer

Ask Claude to output the extracted data in JSON format and include a 'notes' field where it can mention any difficulties encountered.

Answer

Fine-tune Claude on a dataset of annotated invoices to improve its accuracy on handwritten notes, then use it without additional checks.

Question 17

A developer is building a Claude-based application that requires deterministic, structured output in JSON format. The model occasionally returns explanations or markdown code fences around the JSON. Which technique should the developer use to enforce the JSON schema reliably?

Accepted Answer

Define a tool with input_schema matching the desired JSON and instruct Claude to use it.. To reliably obtain structured JSON output from Claude, the most robust method is to define a tool with an input schema that matches the desired JSON structure and instruct Claude to use that tool. This leverages Claude's tool-use capabilities to constrain the output format, ensuring valid JSON that conforms to the schema without extraneous text.

Answer

Increase the temperature setting to 1.0.

Answer

Prepend the prompt with 'You must output only JSON' and set max_tokens to 100.

Answer

Use the 'response_format' parameter with a JSON schema in the request.

Question 18

An engineering team is designing a RAG system using Claude 3.5 Sonnet. They need to ensure the model focuses exclusively on the provided context without hallucinating external knowledge. Which architectural approach best ensures high adherence to provided context?

Accepted Answer

Use XML tags to structure the input and define strict instructions in the system prompt.. Grounding models in provided context requires clear instructions via system prompts and specific document delimiting. By using XML tags to isolate the context and providing explicit constraints in the system prompt, you define the boundaries of the model's knowledge. This architectural pattern is crucial for enterprise applications where accuracy is prioritized over creative generation, minimizing the risk of the model relying on its internal pre-training data.

Answer

Increase the temperature setting to 1.0 to ensure maximum creativity.

Answer

Rely solely on the model's internal training weights for domain-specific queries.

Answer

Reduce the maximum token limit to prevent the model from generating long, incorrect answers.

Question 19

Which capability is a primary benefit of using Claude 3.5 Sonnet compared to smaller, legacy models when processing complex, multi-step instructions?

Accepted Answer

It maintains higher instruction-following performance on complex, multi-step tasks.. Claude 3.5 Sonnet exhibits superior steerability and logical reasoning, allowing it to maintain consistency across long, multi-step prompts. This capability is vital for complex workflows where instructions are layered. Understanding model evolution helps developers choose the right tool for tasks that require high-level reasoning and instruction following, ensuring that complex business logic remains intact throughout the interaction, reducing the need for iterative prompting and debugging.

Answer

It can be trained locally on customer-provided hardware.

Answer

It allows for unlimited parallel API requests without rate limiting.

Answer

It completely removes the need for systematic prompt engineering.

Question 20

Refer to the exhibit. An application sends the provided JSON payload to the Anthropic API. What is the expected behavior regarding the system prompt?

Accepted Answer

The API will correctly apply the system prompt to the entire conversation turn.. In the Anthropic Messages API, the system prompt must be defined at the top level of the JSON object, not within the messages array. Because the system prompt is defined correctly at the top level in this exhibit, it will be correctly processed. Understanding this structure is essential for developers to correctly separate behavioral instructions from conversational history, preventing unexpected model behavior during multi-turn API interactions.

Answer

The API will return an error because the system prompt is improperly formatted.

Answer

The system prompt will be ignored because it must be inside the messages array.

Answer

The model will see the system prompt as the latest user message.