ExamDumpster

GCP-PMLE

Professional Machine Learning Engineer practice test

376 Google GCP-PMLE and GCP-PMLE-Ext practice questions with explanations, in study, timed and flashcard modes. 20 sample questions are free.

Full access

$99

  • All 376 questions across GCP-PMLE and GCP-PMLE-Ext, with explanations
  • Study, timed and flashcard modes
  • Progress saved, resume any time
Try the free samples first
Google logo
Total questions
376
Free samples
20
Versions
2
Updated
Jul 2026

Exam simulator

Practice GCP-PMLE in the simulator.

Study, timed and flashcard modes. Answers and explanations as you go, and your progress saves.

At a glance

GCP-PMLE, by the numbers.

Official specifications published by Google.

Duration
120 min
Questions
50-60
Passing score
70%
Exam fee
$200
In the bank
GCP-PMLE (241), GCP-PMLE-Ext (135)
Open the full exam guide

Exam guide

Everything about GCP-PMLE.

GCP-PMLE · Google · v3.1

What is GCP-PMLE?

GCP-PMLE is a professional Google Cloud exam on machine learning (ML) for engineers who put models into production. Candidates are usually ML engineers or data scientists who code in Python and SQL. It is less about the math behind models and more about running them well on Google Cloud. A large part covers ML pipelines, the automated steps that prepare data, train a model and deploy it. For example, you may decide when a model should be retrained automatically because its accuracy has dropped. This practice test covers low-code AI, model training, serving, pipelines and monitoring.

What you'll practise

  • Build models with BigQuery ML, AutoML, ML APIs and foundation models.
  • Explore and prepare data, prototype in Jupyter notebooks and track experiments.
  • Train models at scale and choose the right hardware for training.
  • Serve models and scale online predictions as traffic grows.
  • Build end-to-end ML pipelines with automated retraining and tracked metadata.
  • Monitor AI solutions for risks, then test and troubleshoot them.

Requirements

  • No formal prerequisites: anyone can book the exam.
  • Recommended: three or more years of industry experience, including a year or more on Google Cloud.
  • Recommended: skill in Python and SQL, and familiarity with TensorFlow, PyTorch or scikit-learn.
  • Recommended: hands-on work with Google Cloud ML tools such as BigQuery ML and AutoML.

Who it's for

  • Machine learning engineers who deploy and maintain models on Google Cloud.
  • Data scientists who want to move models from notebooks into production.
  • MLOps and data engineers who build training and serving pipelines.

Exam details

Published by Google, plus what is in our question bank.

Duration
120 min
Questions
50-60
Passing score
70%
Level
Professional
Question types
Multiple Choice, Multiple Select

Three ways to practise

Study mode

Work through the bank at your own pace with the answer and explanation one tap away. Flag questions, add notes, bookmark the hard ones.

Open in the simulator

Timed mode

A shuffled deck with a countdown that mirrors the real exam. Your score is saved to your attempt history.

Open in the simulator

Flashcards

Question on the front, answer on the back. Good for the last week before the exam.

Open in the simulator

GCP-PMLE questions, answered