Amazon AWS Certified Generative AI Developer - Professional AIP-C01 Practice Test 26

Question 250

A client claims a successful transfer after model generated tool intent. Which specific action is required by this Bedrock API behavior?

Display a success toast immediately after the model issues a transfer tool request.
Treat a queued asynchronous transfer as settled, with no pending state.
Wait for authoritative transaction response before asserting completion.
Generate a plausible receipt number from customer details instead of querying the ledger.

Explanation

Correct Answer

A transfer is complete only after the financial backend confirms commit and returns an authoritative transaction reference.

Incorrect Answers

Tool intent does not prove funds were moved.

Queue acceptance differs from final bank transaction status.

A fabricated identifier is not a verifiable commit receipt.

Question 251

An inference request returns throttling. What should the integration inspect or configure at invocation time?

Retry the request as fast as possible until Bedrock accepts it.
Switch to a model ID not confirmed available in the Region just to avoid throttling.
Implement bounded retry with jitter and quota-aware queue or fallback.
Ignore 429/ThrottlingException and return the model's previously generated answer as a new response.

Explanation

Correct Answer

Use bounded exponential backoff with jitter and an admission queue or validated fallback based on Bedrock throttling and quotas.

Incorrect Answers

A retry storm compounds throttling and can exceed budgets.

A guessed fallback may not be supported or compliant.

A failure is not proof the new request was processed.

Question 252

A client retries a non-idempotent downstream tool after timeout. Which API-level change or validation correctly resolves the condition?

Execute the operation again with a brand-new transaction ID without checking the first attempt.
Query committed transaction status and reuse stable request identity.
Use the model's stated confidence that the first attempt failed as the retry criterion.
Tell the user the operation succeeded because the HTTP client timed out after writing the request body.

Explanation

Correct Answer

On a timeout after a potentially non-idempotent tool call, inspect the authoritative operation state and reuse its original idempotency key.

Incorrect Answers

The initial request may already have committed, creating duplicate side effects.

Model confidence cannot establish the backend's transaction status.

Transport completion is not authoritative commit evidence.

Question 253

A token-heavy request is rejected for exceeding context. Which specific action is required by this Bedrock API behavior?

Increase maxTokens beyond the model's supported context capacity.
Allocate input/output token limits and truncate only with explicit evidence handling.
Truncate the beginning of every source document silently until the call succeeds.
Switch the answer format to JSON without reducing the oversized context.

Explanation

Correct Answer

Budget input plus reserved output tokens for the exact model limit; reduce context with source-aware retrieval before retrying.

Incorrect Answers

Requesting an impossible limit cannot admit an oversized prompt.

Blind trimming can remove governing clauses and invalidate evidence.

Response syntax does not create additional available input tokens.

Question 254

A model sends a JSON syntax error despite configured prompt. What should the integration inspect or configure at invocation time?

Accept an invalid JSON reply because the model was instructed to output JSON.
Validate response and trigger a bounded repair or human exception path.
Use regular expressions to strip all braces until the output parses.
Retry indefinitely with higher temperature until valid syntax appears.

Explanation

Correct Answer

Parse, schema-validate and semantically validate model JSON; bounded repair or explicit exception is safer than accepting malformed data.

Incorrect Answers

Prompt instructions are not a parser or schema validator.

Destructive textual surgery can silently change field meanings.

Unbounded sampling is neither reliable nor cost controlled.

Question 255

A mobile app sends AWS credentials directly with model call. Which API-level change or validation correctly resolves the condition?

Embed an AWS access key in JavaScript and rely on obscured variable names.
Give the frontend a full Bedrock administrator policy to avoid backend token validation.
Use an output content filter as the sole safeguard for a directly exposed runtime endpoint.
Route requests through an authenticated backend rather than exposing cloud credentials.

Explanation

Correct Answer

Do model calls through an authenticated server that uses IAM roles/short-lived credentials, not exposed secrets in the mobile bundle.

Incorrect Answers

Mobile and browser client code can be inspected by attackers.

Broad cloud privileges violate least privilege and caller accountability.

Moderation does not secure AWS credentials or request identities.

Question 256

An SDK adapter incorrectly sends ListFoundationModels through bedrock-runtime and sends Converse through the bedrock control-plane client. Which correction aligns both operations with the proper AWS service endpoints?

Use the bedrock control-plane client for model catalog/configuration APIs and the bedrock-runtime client for Converse or InvokeModel inference.
Call ListFoundationModels on bedrock-runtime using the Converse response schema.
Send InvokeModel requests through the bedrock control client because it has model permissions.
Use the bedrock-agent control API as a universal replacement for Bedrock Runtime inference.

Explanation

Correct Answer

Model catalog and administrative operations belong to the Bedrock control plane, while Converse and InvokeModel are runtime operations. SDK service clients are not interchangeable merely because both mention Bedrock.

Incorrect Answers

Metadata operations and inference calls use different service clients/operations.

Permissions do not change the endpoint that implements the operation.

Agent management is not the same service operation as foundation-model generation.

Question 257

An API Gateway integration exceeds its timeout during long generation. What should the integration inspect or configure at invocation time?

Set an API Gateway timeout to an undocumented infinite value.
Mark HTTP 200 completed as soon as the inference request is queued.
Design streaming or asynchronous continuation under actual gateway limits.
Increase temperature so the model supposedly returns faster.

Explanation

Correct Answer

When the gateway cannot hold a long generation within its supported timeout, use a supported stream or durable async job with status contract.

Incorrect Answers

Managed gateway integrations have service-specific limits.

Asynchronous acceptance is not final completion.

Sampling variation does not reliably remove an integration timeout.

Question 258

An SSE client loses a stream halfway through output. Which API-level change or validation correctly resolves the condition?

Render the partial stream as a final verified answer after transport closes unexpectedly.
Report incomplete response and recover from last safe checkpoint, not declare success.
Concatenate the stream with a model-generated imagined ending and mark it complete.
Retry the entire non-idempotent downstream action with a new ID after a stream break.

Explanation

Correct Answer

Handle stream cancellation and errors distinctly; incomplete SSE content must remain partial and allow an explicit safe retry/recovery path.

Incorrect Answers

Missing tokens may change the claim or omit required fields.

An inferred continuation is not an executed response from the original request.

Transport loss does not establish side-effect failure.

Question 259

A request is rejected due to unknown model ID. Which specific action is required by this Bedrock API behavior?

Assume a provider-family marketing name is a valid runtime modelId.
Replace the model identifier with an arbitrary old alias taken from a different Region.
Grant S3 wildcard permissions instead of checking model ID availability.
Resolve exact current model identifier and Region via ListFoundationModels/model card.

Explanation

Correct Answer

Resolve the exact current model ID and its Region via model listing/cards and verify any required inference profile ARN.

Incorrect Answers

An API identifier must match an available model or deployment resource.

Identifiers and lifecycle support differ by Region and version.

Object permissions do not create the requested model endpoint.

Question 260

A model response reaches the configured output-token limit while generating a structured invoice update. The partial object is syntactically incomplete and the application is about to apply it to an authoritative finance record. What commit-time gate must the caller enforce?

Accept the partial JSON prefix as valid because it begins with a curly brace.
Reject truncated or incomplete output based on stop reason and schema/business validation before any financial write, and retry within supported output limits only when safe.
Assume the model can always resume the exact token stream if a second request is issued.
Increase network timeout only while leaving max output token cap unchanged.

Explanation

Correct Answer

A max-token stop can leave a partial JSON or business command. An incomplete model response must not become an authoritative transaction merely because it contains plausible fields.

Incorrect Answers

The response may be syntactically/semantically incomplete.

A separate generation is not guaranteed to be an exact continuation.

Time limits do not repair a generation stopped by token exhaustion.

Pass Your Amazon AWS Certified Generative AI Developer - Professional AIP-C01 Exam Easy!

Question 250

Explanation

Correct Answer

Incorrect Answers

Question 251

Explanation

Correct Answer

Incorrect Answers

Question 252

Explanation

Correct Answer

Incorrect Answers

Question 253

Explanation

Correct Answer

Incorrect Answers

Question 254

Explanation

Correct Answer

Incorrect Answers

Question 255

Explanation

Correct Answer

Incorrect Answers

Question 256

Explanation

Correct Answer

Incorrect Answers

Question 257

Explanation

Correct Answer

Incorrect Answers

Question 258

Explanation

Correct Answer

Incorrect Answers

Question 259

Explanation

Correct Answer

Incorrect Answers

Question 260

Explanation

Correct Answer

Incorrect Answers

AWS Certified Generative AI Developer - Professional AIP-C01 Practice Tests