Learn mode
Learn as you go
Choose an answer, reveal the marked key, and move through the bank at your own pace.
364 questions open free
Start guided practiceGoogle Professional Machine Learning Engineer
Production machine learning on Google Cloud: framing problems, building and scaling models on Vertex AI, serving them reliably, and automating and monitoring the pipeline.
Google Cloud does not publish numeric exam codes — PMLE is the common abbreviation.
Google Professional Machine Learning Engineer is Google Cloud's professional-level ai & machine learning certification, referred to across the industry by the shorthand PMLE. The exam runs for 2 hours and presents up to 55 questions, with a pass mark of not published by Google. It costs US$200 at the standard rate.
The published exam guide lists 6 sections but does not assign public weightings, so plan to cover each in similar depth rather than betting on one being dominant.
Confirm these details on the official Google Cloud page before booking — vendors adjust format and pricing without notice.
Some forms include unscored trial questions
Standard rate, before local taxes
The published exam guide for PMLE, in the vendor's own order.
01Architecting low-code AI solutions
02Collaborating within and across teams to manage data and models
03Scaling prototypes into machine learning models
04Serving and scaling models
05Automating and orchestrating machine learning pipelines
06Monitoring AI solutions
Three or more years of experience including at least one year designing ML solutions on Google Cloud.
Professional level — Advanced credentials covering design, scale and trade-off decisions.
Start with 364 complete questions in the PMLE bank. Free progress and bookmarks stay in this browser.
Choose an answer, reveal the marked key, and move through the bank at your own pace.
364 questions open free
Start guided practiceKeep answers hidden on the clock and review every miss after you finish the public set.
Free timed practice
Open timed mockAll questions and practice tests are free.
No sign-up, payment or access expiry. Your progress stays in this browser.
Questions, answered
Clear answers about the exam, the study experience, and what to expect before you begin.
Google Cloud lists up to 55 questions, to be completed in 2 hours. Question counts can vary slightly between exam forms, and unscored trial questions are sometimes included.
Not published by Google. Confirm the current figure on the official exam page before booking, as vendors adjust cut scores when an exam form is refreshed.
The standard price is US$200. Local pricing, taxes and any retake policy vary by region, so check the official page for the figure that applies to you.
Three or more years of experience including at least one year designing ML solutions on Google Cloud.
The exam guide covers Architecting low-code AI solutions, Collaborating within and across teams to manage data and models, Scaling prototypes into machine learning models, Serving and scaling models and Automating and orchestrating machine learning pipelines, plus 1 further domain. The full weighted breakdown is listed above.
Yes — this exam has 364 practice questions available here, each with the correct answer marked.