Generative AI Leader Google Cloud's Generative AI Offerings Practice Test 11 — 15 Questions

Question 1

A media company needs to generate short video clips from text prompts for a marketing campaign. They want to use a Google Cloud service that can create videos up to 8 seconds long, 720p resolution, and can be accessed via Vertex AI. Which Google Cloud generative AI offering should they use?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Veo on Vertex AI. Veo on Vertex AI is specifically built for video generation from text, supporting up to 8-second 720p clips, which aligns with the media company's need for short marketing videos. Other models like Imagen, Gemini, or Chirp serve different modalities and cannot produce video outputs.

Answer

Imagen on Vertex AI

Answer

Gemini on Vertex AI

Answer

Chirp on Vertex AI

Question 2

A startup wants to quickly prototype a generative AI application that can write marketing copy. They have limited machine learning expertise and want to avoid managing infrastructure. They prefer a fully managed, no-code or low-code solution that provides access to Google's foundation models. Which Google Cloud offering should they use?

Accepted Answer

Vertex AI Studio. Vertex AI Studio is the correct choice because it offers a user-friendly, no-code environment for experimenting with and prototyping generative AI models. It allows users to design prompts, test models like Gemini, and even fine-tune them without managing infrastructure. The other options are either too technical or not focused on generative AI prototyping.

Answer

BigQuery ML

Answer

Vertex AI Pipelines

Answer

Cloud Functions

Question 3

A software company wants to use a Google Cloud generative AI model to automatically generate code snippets from natural language descriptions. The developers need the model to produce syntactically correct code in multiple programming languages. Which Google Cloud offering is specifically designed for this use case?

Accepted Answer

Codey APIs on Vertex AI. Codey APIs on Vertex AI are purpose-built for code generation tasks, including translating natural language to code in multiple languages. They are trained on a large corpus of code and are optimized for syntax and correctness. The other options are either general multimodal models, search services, or conversational agents, none of which are specialized for code generation.

Answer

Dialogflow CX

Answer

Vertex AI Matching Engine

Answer

Gemini Pro Vision

Question 4

An enterprise architect must let a fleet of internal applications call Gemini models on Google Cloud while enforcing VPC Service Controls perimeters, customer-managed encryption keys, and audit logging tied to the organization's existing IAM. Which Google Cloud offering should the architect standardize on?

Accepted Answer

Vertex AI with Gemini models. Vertex AI is the Google Cloud platform where Gemini models are served to applications under the same enterprise controls as other Cloud services, including IAM, VPC Service Controls, customer-managed encryption keys, and audit logging. Because the architect needs perimeter enforcement and key management for API traffic, the Vertex AI model endpoints are the appropriate foundation rather than consumer-style APIs or assistant products.

Answer

Gemini API via Google AI Studio

Answer

Gemini Enterprise with data connectors

Answer

Gemini for Google Workspace

Question 5

A logistics company wants to build a generative AI application that answers questions over thousands of internal policy PDFs stored in Cloud Storage. They need Google Cloud to handle document ingestion, chunking, indexing, and retrieval for grounding, while they focus only on the application logic. Which Google Cloud offering should they use?

Accepted Answer

Vertex AI Search. Vertex AI Search is purpose-built for ingesting enterprise documents, chunking and indexing them, and retrieving relevant passages to ground generative responses. It supports Cloud Storage as a data source and abstracts away the retrieval pipeline, so the logistics team can focus on application logic. The other services handle features, document parsing, or image analysis and do not deliver managed semantic retrieval for grounding.

Answer

Cloud Vision API

Answer

Vertex AI Feature Store

Answer

Document AI

Question 6

A bank is deploying a Gemini model on Vertex AI to draft responses to customer complaints. Compliance requires that the deployment (Choose two.)

Accepted Answer

Records prompts and responses for audit purposes. A compliant generative AI deployment in a regulated setting needs traceability and output control. Logging prompts and responses creates the audit trail examiners expect, while configurable safety filters prevent harmful or policy-violating content from reaching customers. Bulk unreviewed training data, disabled logging, and public weight release either increase risk or remove evidence, so they do not belong in this deployment.

Answer

Disables all logging to reduce storage costs

Answer

Trains the model on the bank's entire historical email archive without review

Answer

Publishes the model's weights publicly to demonstrate transparency

Question 7

A healthcare provider is building a generative AI assistant on Vertex AI to answer patient questions about appointment scheduling. They must ensure the model does not produce harmful or medically unsafe content, and they want to monitor and tune safety behavior over time. Which two Google Cloud features should they use? (Choose two.)

Accepted Answer

Vertex AI Model Monitoring for skew and drift detection. Safety filters with configurable thresholds block harmful or medically unsafe inputs and outputs, directly protecting patients. Model Monitoring tracks changes in data and predictions over time, surfacing drift that may indicate new safety risks or emerging question patterns. Together they provide both real-time protection and ongoing oversight, which the healthcare provider requires for a patient-facing assistant.

Answer

BigQuery ML logistic regression for sentiment analysis

Answer

Cloud KMS customer-managed encryption keys

Answer

Cloud Trace for distributed latency tracing

Question 8

A university's IT department is evaluating Google Cloud generative AI offerings to build a course-assistant tool for students. They want to reduce engineering effort by using managed services rather than hosting models themselves. Which two Google Cloud offerings should they consider? (Choose two.)

Accepted Answer

Vertex AI Agent Builder for assembling a grounded agent with tools and a knowledge base.. The department wants managed services to cut engineering effort. Gemini through Vertex AI supplies the model without infrastructure ownership, and Agent Builder supplies the orchestration, grounding, and tooling layer needed for a usable course assistant. Self-managing a model on GKE, pretraining a custom model, or forwarding requests to an external API all reintroduce significant engineering or governance work that the managed first-party offerings avoid.

Answer

A custom training pipeline built on Vertex AI Training to pretrain a domain-specific language model.

Answer

A self-managed open-weights model served from a Google Kubernetes Engine cluster with GPU node pools.

Answer

BigQuery ML with a remote model that forwards prediction requests to an external vendor's API.

Question 9

A startup wants to build a conversational AI assistant that can understand and generate text, images, and code. They need a single model that can handle multimodal inputs and outputs. Which Google Cloud generative AI model should they choose?

Accepted Answer

Gemini. Gemini is the only Google Cloud generative AI model designed to be natively multimodal, capable of understanding and generating text, images, and code. This makes it ideal for a conversational AI assistant that needs to handle diverse content types within a single model.

Answer

Imagen

Answer

Chirp

Answer

Codey

Question 10

A large enterprise is using Vertex AI to deploy a generative AI model for internal document summarization. They need to ensure that the model's responses are based on the most current internal documents and that the model does not hallucinate. They also want to minimize latency and cost. Which feature of Vertex AI should they implement?

Accepted Answer

Grounding with Vertex AI Search. Grounding with Vertex AI Search is the correct feature because it enables the model to retrieve and cite relevant passages from a designated data store, ensuring responses are based on the latest internal documents. This approach reduces hallucinations and avoids the need for frequent fine-tuning, which can be costly and slow. It also supports low-latency responses by integrating retrieval with generation.

Answer

Model evaluation

Answer

Batch prediction

Answer

Model fine-tuning

Question 11

A financial services company wants to use generative AI to summarize large volumes of internal documents while ensuring that sensitive data never leaves their virtual private cloud (VPC). They need a solution that provides Gemini models with enterprise-grade security and data residency controls. Which Google Cloud service should they use?

Accepted Answer

Vertex AI with Gemini models. Vertex AI offers Gemini models with enterprise-grade security, including VPC Service Controls and data residency, which are critical for handling sensitive financial data. It enables the company to use generative AI while maintaining strict control over data location and access. The other services either lack generative capabilities or do not provide the necessary security and compliance features for this scenario.

Answer

Gemini API in Google AI Studio

Answer

Document AI

Answer

Cloud Natural Language API

Question 12

A media production studio wants to create short video clips from text prompts for social media content. They need a Google Cloud generative AI model that can generate videos from text. Which offering should they use?

Accepted Answer

Veo. Veo is the Google Cloud generative AI model specifically built for text-to-video generation. It enables users to create short video clips from textual descriptions, which aligns perfectly with the media studio's goal of producing social media content. Other models like Imagen and Gemini handle images or text but not video output, so Veo is the correct offering.

Answer

Chirp

Answer

Gemini

Answer

Imagen

Question 13

A global logistics company wants to build a generative AI assistant that can answer operational questions by retrieving information from internal PDF and HTML documents stored in Cloud Storage. They need a managed, serverless retrieval-augmented generation (RAG) capability that requires minimal infrastructure management and integrates with Vertex AI. Which Google Cloud service should they use?

Accepted Answer

Vertex AI Search. Vertex AI Search delivers a fully managed, serverless retrieval-augmented generation pipeline that ingests documents from Cloud Storage, builds an index, and grounds Gemini responses without requiring the team to manage embeddings or vector databases. The other services are either feature stores, self-managed vector databases, or analytics tools that do not provide end-to-end RAG for unstructured content.

Answer

Vertex AI Feature Store

Answer

Cloud SQL with pgvector

Answer

BigQuery ML

Question 14

A media company wants to let its editors query a large archive of internal video transcripts using everyday conversational questions, and the app must return grounded answers that cite the exact source clips. The team has no machine learning engineers and wants the least operational overhead. Which Google Cloud offering should they use?

Accepted Answer

Vertex AI Search with a connected transcript data store. A managed search-and-grounding service is the right fit when business users need conversational answers over an indexed corpus with citations and no ML engineering effort. Indexing transcripts into a data store and letting the search service handle retrieval, grounding and citation generation satisfies both the accuracy and low-overhead constraints. Custom pipelines or raw model calls push that work back onto the team.

Answer

BigQuery ML with a remote Gemini model reference

Answer

Vertex AI Pipelines with a custom retrieval component

Answer

Gemini via the Gemini API in a stateless prompt loop

Question 15

A logistics company wants employees to ask natural-language questions about shipment trends and have Gemini generate SQL, charts, and narrative summaries directly against data already stored in BigQuery, without moving or duplicating that data. Which Google Cloud capability should the company adopt?

Accepted Answer

Gemini in BigQuery. Gemini in BigQuery is designed to help users write SQL, understand results, and produce summaries and visualizations using data that remains in BigQuery. Since the logistics team wants natural-language insight without duplicating datasets, the embedded BigQuery assistant is the direct fit rather than exports, external deployments, or workspace-side copies.

Answer

Vertex AI Model Garden with a deployed open model

Answer

Gemini for Google Workspace connected to a BigQuery export sheet

Answer

Exporting BigQuery tables to Cloud Storage, then querying them with a custom Gemini application