Claims register

Each section is reported on a 1–6 scale in half-point steps, effective 21 January 2026.

The overall score is the average of the four section scores, rounded to the nearest half point.

Through January 2028, score reports also carry a CEFR level and a comparable total on the retired 0–120 scale.

ETS machine-scores 108 items and AI-scores the other 12.

Reading and Listening include non-scored items used to trial future content. Reading carries 35 raw points across 50 items and Listening 35 across 47, so 27 of the 120 items on the test do not count toward a score.

Maximum points per item: 1 for scored Reading and Listening items, 1 for Build a Sentence, 5 for Write an Email, 5 for Write for an Academic Discussion, 5 for Listen and Repeat, 5 for Take an Interview.

AI scoring applies to constructed responses in Speaking and Writing and assesses features such as fluency, coherence and grammar. Machine scoring applies rule-based logic to selected-response items.

120 items in total, carrying 145 raw points.

Total test time is between 1 hr 23 min and 1 hr 29 min, depending on adaptive routing.

The order the four sections are administered in.

50 items. A router module of 18–21 minutes sets the difficulty of a 9-minute second module. All items machine scored.

Task types: Complete the Words (30 items, B1–C1+), Read an Academic Passage (5–15, B1–C2), Read in Daily Life (5–15, A1–C1).

Easier texts run roughly 15–50 words and more complex texts up to 200. Complexity is calibrated with TextEvaluator: medium 800–1100, higher 1100–1360+. Flesch-Kincaid: medium 7.4–10.5, high 10.5–20.0. All texts stay visible throughout the task.

Stimuli range from short everyday texts such as signs, menus, emails and social media posts to extended academic passages. Read in Daily Life is delivered in 2-item and 3-item sets.

47 items. An 18-minute router module, then a second module of 7 or 11 minutes. All items machine scored.

Task types: Listen and Choose a Response (15–19 items, A1–B2), Listen to a Conversation (10, A2–C1), Listen to an Academic Talk (8–16, A2–C2), Listen to an Announcement (6–10, A2–C1).

Audio ranges from brief utterances of no more than six stressed syllables to extended monologue of up to 250 words, with intermediate input typically 35–100 words. Recordings play once. Each stimulus is accompanied by a visual of the speaker.

Listening stimuli use AI-generated audio and sometimes human voice actors, with a generally balanced representation of accents including US/Canadian, Australian and British English.

12 items in 23 minutes: Build a Sentence (10, A1–C2, machine scored), Write an Email (1, B1–C2, AI scored), Write for an Academic Discussion (1, B1–C2, AI scored).

Build a Sentence accounts for 10 of the Writing section's 20 raw points; the two constructed responses account for the other 10.

11 items in 8 minutes: Listen and Repeat (7, A1–C2) and Take an Interview (4, A1–C2). Both AI scored.

Listen and Repeat accounts for 35 of the Speaking section's 55 raw points; Take an Interview accounts for 20.

Each Speaking task opens with a scenario introduction delivered both aurally and in print, with visual context such as schematic images, maps, forms or photos of speakers.

Scores are available in roughly 72 hours.

Reading and Listening are two-stage adaptive, timed as a router plus a lower or upper module. Writing and Speaking are linear: every test taker on a given form receives the same tasks.

The specifications document states that it is subject to minor revisions until the assessment's official launch, and carries no visible version number.

98 questions in total: 54 Reading and Writing, 44 Math.

2 hr 14 min of testing time, plus a 10-minute break between sections.

Two 32-minute Reading and Writing modules of 27 questions each, and two 35-minute Math modules of 22 questions each.

8 questions are unscored pretest items, two per module, so 90 of the 98 are scored.

Total scale 400–1600, with each section reported on 200–800.

Everyone sits the same mixed-difficulty Module 1. Performance there routes you to a higher- or lower-difficulty Module 2, which affects the ceiling on that section score.

Math is roughly three quarters multiple choice and one quarter student-produced response.

Passages are short, roughly 25–150 words, and each carries a single question.

Reading and Writing: Information and Ideas, Craft and Structure, Expression of Ideas, Standard English Conventions. Math: Algebra, Advanced Math, Problem-Solving and Data Analysis, Geometry and Trigonometry.

English delivers 50 items, scores 40, and takes 35 minutes.

Mathematics delivers 45 items, scores 41, and takes 50 minutes.

Reading delivers 36 items, scores 27, and takes 40 minutes.

A calculator is permitted on the mathematics test only.

Each ACT test section generates a single scale score between 1 and 36.

The composite is the mean of the English, Mathematics, and Reading scale scores. A fractional mean rounds to the nearest whole number: a fraction less than one-half rounds down, except an average lower than 1 which rounds up; a fraction of one-half or more rounds up.

Section scores are based only on the number of questions answered correctly. There is no penalty for guessing.

Delivered counts include embedded field-test items. Those items are not labeled and do not contribute to the score.

English reporting categories, as a percent of scored items: Production of Writing 38–43, Knowledge of Language 18–23, Conventions of Standard English 38–43.

Mathematics reporting categories, as a percent of the section: Preparing for Higher Math 80, including Number and Quantity 10–12, Algebra 17–20, Functions 17–20, Geometry 17–20, and Statistics and Probability 12–15; Integrating Essential Skills 20. Modeling is counted inside the other categories and is not its own share.

Reading reporting categories, as a percent of scored items: Key Ideas and Details 44–52, Craft and Structure 26–33, Integration of Knowledge and Ideas 19–26.

The target on each reporting category is Acedify's assembly allocation inside the published range. It is not an ACT published point value.

Whether registration requires membership of the Claude Partner Network, or a work email on a partner organisation's domain.

Four credentials: Claude Certified Associate Foundations (CCAO-F), Claude Certified Developer Foundations (CCDV-F), Claude Certified Architect Foundations (CCAR-F), Claude Certified Architect Professional (CCAR-P).

Credentials are valid for 12 months.

All four exams run 120 minutes.

Question count per exam.

Delivered by Pearson VUE, online proctored or at a test centre.

A pass mark of 720 on a 100–1000 scale.

Exam fees per credential.

Domain names and weightings for each of the four exams.

Technical exam scores are reported on a 1 to 1,000 scale, and 700 or greater is a pass. It is a scaled score and does not correspond to 70% of points.

There is no penalty for a wrong answer, and some questions are worth more than one point.

AZ-104 is a Microsoft technical exam: 700 or greater on a 1 to 1,000 scale passes. Pass/fail is the overall score only; the score report's skill-area bars are diagnostic and do not each have a pass mark.

Microsoft states that most certification exams contain between 40 and 60 questions, that the number varies by exam, and that it is subject to change.

Which specific exams include performance-based lab questions.

Duration, price and skills-outline weightings for each individual exam.

Candidates have 120 minutes to complete the AI-103 assessment.

AI-103 is scheduled through Pearson VUE.

Standard list price in USD for AI-103.

Most questions cover features that are generally available. The exam may contain questions on preview features if those features are commonly used.

Candidates have 120 minutes to complete the AI-200 assessment.

AI-200 is scheduled through Pearson VUE.

Standard list price in USD for AI-200.

Question formats include multiple choice, multiple response, drag-and-drop and case studies.

Candidates have 120 minutes to complete the AI-300 assessment.

AI-300 is scheduled through Pearson VUE.

Fixed question count for AI-300.

Exam length for SC-500.

Fixed question count for SC-500.

SC-500 is scheduled through Pearson VUE.

Most questions cover features that are generally available. The exam may contain questions on preview features if those features are commonly used.

AI-901 is offered in English, Arabic (Saudi Arabia), Chinese (Simplified), Chinese (Traditional), French, German, Indonesian (Indonesia), Italian, Japanese, Korean, Portuguese (Brazil), Russian, and Spanish.

Sitting length for AI-901.

Standard list price in USD for AI-901.

AI-901 is scheduled through Pearson VUE.

Candidates have 45 minutes to complete the AZ-900 assessment.

AZ-900 is scheduled through Pearson VUE. Students and educators can also schedule through Certiport.

AZ-900 is offered in English, Japanese, Chinese (Simplified), Korean, Spanish, German, French, Indonesian (Indonesia), Arabic (Saudi Arabia), Chinese (Traditional), Italian, Portuguese (Brazil), and Russian.

Standard list price in USD for AZ-900.

Candidates have 100 minutes to complete the AZ-104 assessment.

AZ-104 is scheduled through Pearson VUE.

Standard list price in USD for AZ-104.

AZ-104 question formats include multiple choice, multi-select, drag-and-drop, hotspot and case studies.

Exam AI-102 was retired on 30 June 2026 at 11:59 PM Central Standard Time. Exam AI-103, Developing AI Apps and Agents on Azure, replaces it.

Exam AI-103 exists, titled Developing AI Apps and Agents on Azure, and is the exam for Microsoft Certified: Azure AI Apps and Agents Developer Associate.

AI-103 beta April 2026, generally available around June 2026.

AI-102 has no renewal path once retired.

Exam AZ-204 was retired on 31 July 2026 at 11:59 PM Central Standard Time. Exam AI-200, Developing AI Cloud Solutions on Azure, replaces it.

Microsoft Certified: Azure Developer Associate and its renewal assessment are retired. There is no renewal path for the old credential.

Exam AI-200 exists, titled Developing AI Cloud Solutions on Azure, and is the exam for Microsoft Certified: Azure AI Cloud Developer Associate.

The credential earned by passing AI-200 is Azure AI Cloud Developer Associate.

AI-200 beta April 2026, full launch July 2026.

Exam AI-300 exists, titled Operationalizing Machine Learning and Generative AI Solutions, and is the exam for Microsoft Certified: Machine Learning Operations Engineer Associate.

The credential earned by passing AI-300 is Machine Learning Operations Engineer Associate.

Microsoft associate, expert, and specialty certifications expire annually and can be renewed by passing a free online assessment on Microsoft Learn.

AI-300 is in beta.

Exam SC-500 exists, titled Implementing End-to-End Security Controls for Cloud and AI Workloads, and is the exam for Microsoft Certified: Cloud and AI Security Engineer Associate.

The credential earned by passing SC-500 is Cloud and AI Security Engineer Associate.

SC-500 replaces AZ-500. AZ-500 retired on 31 August 2026.

The AZ-500 renewal assessment retires with the exam, so anyone whose credential expires after the retirement date never gets a renewal window.

SC-500 beta May 2026, GA 21 July 2026.

Exam AI-900 was retired on 30 June 2026 at 11:59 PM Central Standard Time. Exam AI-901, Microsoft Azure AI Fundamentals, replaces it.

Exam AI-901 exists, titled Microsoft Azure AI Fundamentals, and is the exam for Microsoft Certified: Azure AI Fundamentals.

Passing AI-901 earns Microsoft Certified: Azure AI Fundamentals.

AI-900 and AI-901 award the same credential, Microsoft Certified: Azure AI Fundamentals.

The English language version of the exam was updated on 15 April 2026. Localized versions are updated approximately eight weeks later.

The English language version of the Azure AI Fundamentals certification was updated on 15 April 2026.

Exam AZ-900 exists, titled Microsoft Azure Fundamentals, and is the exam for Microsoft Certified: Azure Fundamentals. The exam page carries no retirement notice.

Passing AZ-900 earns Microsoft Certified: Azure Fundamentals.

Exam AZ-104 exists, titled Microsoft Azure Administrator, and is the exam for Microsoft Certified: Azure Administrator Associate. The exam page carries no retirement notice.

Passing AZ-104 earns Microsoft Certified: Azure Administrator Associate.

Azure Administrator Associate renews every 12 months, at no cost, by passing an online assessment on Microsoft Learn.

Exam DP-100 was retired on 1 June 2026 at 11:59 PM Central Standard Time.

AI-300 replaces DP-100.

Exam AZ-500 was retired on 31 August 2026 at 11:59 PM Central Standard Time.

SC-500 replaces AZ-500.

Microsoft fundamentals certifications do not expire.

Microsoft's audience profile for AI-103 expects experience developing apps with Python, plus familiarity with general AI, generative AI and Azure services.

Microsoft's audience profile for AI-200 covers all phases of implementing AI solutions on Azure, with an emphasis on back-end services and components, across requirements, design, development, deployment, security and monitoring. Python is listed as a proficiency.

Microsoft's audience profile for AI-300 covers MLOps and GenAIOps on Azure using Azure Machine Learning, Microsoft Foundry, GitHub Actions, Bicep, Azure CLI, and Python.

Microsoft's audience profile for SC-500 assumes practical administration of Azure and hybrid environments, strong familiarity with Microsoft Entra ID, and familiarity with Microsoft 365 administration.

Microsoft's audience profile for AI-901 is a candidate at the beginning of a career in AI solution development, with conceptual knowledge of AI solutions in Azure, foundational technical skills, Python coding syntax and programming techniques, and familiarity with Azure resources. A study-guide note also expects familiarity with REST APIs, SDKs, and CLIs.

AZ-900 has no prerequisite exams or credentials.

AZ-104 has no prerequisite exams. The study guide lists expected experience, not a required prior credential. AZ-900 is not named as required.

The candidate profile expects subject-matter expertise implementing, managing and monitoring an Azure environment, plus experience with PowerShell, Azure CLI, the portal, ARM templates or Bicep, and Microsoft Entra ID.

Skills measured as of 16 April 2026.

Five domains: Plan and manage an Azure AI solution (25–30%); Implement generative AI and agentic solutions (30–35%); Implement computer vision solutions (10–15%); Implement text analysis solutions (10–15%); Implement information extraction solutions (10–15%).

Skills measured as of 15 April 2026.

Four domains: Develop containerized solutions on Azure (20–25%); Develop AI solutions by using Azure data management services (25–30%); Connect to and consume Azure services (20–25%); Secure, monitor, and troubleshoot Azure solutions (20–25%).

Five domains: Design and implement an MLOps infrastructure (15–20%); Implement machine learning model lifecycle and operations (25–30%); Design and implement a GenAIOps infrastructure (20–25%); Implement generative AI quality assurance and observability (10–15%); Optimize generative AI systems and model performance (10–15%).

The AI-300 study guide was last updated on 5 March 2026.

Skills measured as of [date] for AI-300.

Four domains: Manage identity, access, and governance (20–25%); Secure storage, databases, and networking (25–30%); Secure compute (20–25%); Manage and monitor security posture (20–25%).

"Implement security for AI" is a sub-section of Secure compute, not a domain of its own.

The AI sub-section names Entra Agent ID, Copilot and DSPM risk, Defender for AI Service, AI Gateway for Microsoft Foundry, and Foundry agent guardrails.

Skills measured as of [date] for SC-500.

Skills measured as of 15 April 2026.

Two domains: Identify AI concepts and capabilities (40–45%); Implement AI solutions by using Microsoft Foundry (55–60%).

Skills measured as of 20 July 2026.

The 20 July 2026 refresh made minor changes to Describe Azure compute and networking services, Describe features and tools for managing and deploying Azure resources, and Describe monitoring tools in Azure. Domain names and weights were unchanged.

Three domains: Describe cloud concepts (25–30%); Describe Azure architecture and services (35–40%); Describe Azure management and governance (30–35%).

Skills measured were revised on 14 January 2026 before the July 2026 refresh.

Skills measured as of 17 April 2026.

Secondary AZ-104 guides disagree about the skills-outline date: some print 17 April 2026, others print 17 September 2026. Microsoft Learn's study guide, read on 14 September 2026, says 17 April 2026. At least one widely repeated date was wrong.

Five domains: Manage Azure identities and governance (20–25%); Implement and manage storage (15–20%); Deploy and manage Azure compute resources (20–25%); Implement and manage virtual networking (15–20%); Monitor and maintain Azure resources (10–15%).

Identity and governance gained weight in the 2026 revision and virtual networking lost some.

The July 2026 skills outline does not name Microsoft Copilot, Microsoft Fabric, Azure OpenAI, or the six Responsible AI principles.

The April 2026 skills outline names Microsoft Entra ID (not Azure Active Directory), ARM templates and Bicep, Azure Container Apps, Azure Bastion, and Network Watcher.

AWS reports exam results as a scaled score from 100 to 1,000. Scaled scoring equates results across exam forms of differing difficulty, so the score is not a percentage of questions answered correctly.

The minimum passing score varies by exam. Cloud Practitioner (CLF-C02) and AI Practitioner (AIF-C01) pass at 700; Solutions Architect Associate (SAA-C03) passes at 720; Solutions Architect Professional (SAP-C02) passes at 750.

AWS exams use a compensatory scoring model: you do not need a passing score in each section, only overall.

Every AWS exam includes unscored questions used to evaluate items for future use. They are not identified on the exam.

CLF-C02 results are a scaled score from 100 to 1,000. The minimum passing score is 700.

CLF-C02 uses a compensatory scoring model: you do not need to pass each section, only the exam overall.

AWS advises caution when interpreting section-level feedback on a CLF-C02 score report.

SAA-C03 results are a scaled score of 100–1,000. The minimum passing score is 720.

The minimum passing score for SAA-C03 is 720 on a 100–1,000 scale.

SAA-C03 uses a compensatory scoring model: you do not need to achieve a passing score in each section, only the exam overall.

The SAA-C03 score report may include a table of classifications of performance at each section level. AWS advises caution when interpreting that section-level feedback.

DVA-C02 results are a scaled score of 100–1,000. The minimum passing score is 720.

DVA-C02 uses a compensatory scoring model: you do not need to pass each section, only the exam overall.

AIF-C01 results are a scaled score of 100–1,000. The minimum passing score is 700.

AIF-C01 uses a compensatory scoring model: you do not need to achieve a passing score in each section, only the exam overall.

Unanswered questions are marked incorrect, so every question should be attempted.

On CLF-C02, unanswered questions are scored as incorrect and there is no penalty for guessing.

On SAA-C03, unanswered questions are scored as incorrect and there is no penalty for guessing.

On AIF-C01, unanswered questions are scored as incorrect and there is no penalty for guessing.

SAA-C03 has 65 questions in 130 minutes, in multiple-choice and multiple-response formats.

CLF-C02 has 65 questions, of which 50 are scored and 15 are unscored. The unscored questions do not affect the score and are not identified on the exam.

CLF-C02 has two question types: multiple choice (one correct of four) and multiple response (two or more correct of five or more).

Exam duration for CLF-C02.

Exam price for CLF-C02.

SAA-C03 has 65 questions, of which 50 affect your score.

SAA-C03 includes 15 unscored questions that do not affect the score and are not identified on the exam.

SAA-C03 exam duration is 130 minutes.

SAA-C03 has two question types: multiple choice (one correct of four) and multiple response (two or more correct of five or more).

Exam price for SAA-C03.

DVA-C02 exam duration is 130 minutes.

DVA-C02 has 65 questions, of which 50 affect your score.

DVA-C02 includes 15 unscored questions that do not affect the score and are not identified on the exam.

DVA-C02 has two question types: multiple choice and multiple response.

AIF-C01 includes 50 questions that affect your score.

AIF-C01 includes 15 unscored questions that do not affect the score and are not identified on the exam.

AIF-C01 has four question types: multiple choice, multiple response, ordering, and matching. Multiple response, ordering, and matching require every correct selection to receive credit.

Exam duration for AIF-C01.

Exam price for AIF-C01.

Delivery method and validity period for AIF-C01.

Current version code and retirement status for every AWS exam listed on the site.

DVA-C03 registration opens on 27 October 2026. AWS publishes DVA-C03 exam details on the certification page the same day.

The last day to sit DVA-C02 is 1 December 2026.

CLF-C02 scored content is weighted Cloud Technology and Services 34%, Security and Compliance 30%, Cloud Concepts 24%, and Billing, Pricing, and Support 12%.

Coding, designing cloud architecture, troubleshooting, implementation, and load and performance testing are out of scope for the CLF-C02 target candidate.

SAA-C03 validates a candidate's ability to design solutions based on the AWS Well-Architected Framework.

SAA-C03 scored content is weighted Design Secure Architectures 30%, Design Resilient Architectures 26%, Design High-Performing Architectures 24%, and Design Cost-Optimized Architectures 20%.

Four domains at 32 / 26 / 24 / 18 of scored content: Development with AWS Services, Security, Deployment, Troubleshooting and Optimization.

AIF-C01 scored content is weighted Fundamentals of AI and ML 20%, Fundamentals of GenAI 24%, Applications of Foundation Models 28%, Guidelines for Responsible AI 14%, and Security, Compliance, and Governance for AI Solutions 14%.

Out of scope for the AIF-C01 target candidate: developing or coding AI/ML models or algorithms; implementing data or feature engineering; hyperparameter tuning or model optimization; building and deploying AI/ML pipelines or infrastructure; mathematical or statistical analysis of models; implementing security or compliance protocols for AI/ML systems; developing and implementing governance frameworks and policies for AI/ML solutions.

The target candidate has up to six months of exposure to AWS Cloud design, implementation, or operations.

The target candidate should have at least 1 year of hands-on experience designing cloud solutions that use AWS services.

The target candidate has 1+ years developing and maintaining applications using AWS services.

The target candidate has up to 6 months of exposure to AI/ML technologies on AWS, and uses but does not necessarily build AI/ML solutions on AWS.

The target candidate should be familiar with core AWS services, the AWS shared responsibility model, IAM, and AWS service pricing models.

The exam code is SAA-C03. The credential is AWS Certified Solutions Architect – Associate.

SAA-C03 is an Associate exam.

The exam code is DVA-C02. The credential is AWS Certified Developer – Associate.

DVA-C02 is an Associate exam.

The exam code is AIF-C01. The credential is AWS Certified AI Practitioner.

AIF-C01 is a Foundational exam.

Paper 1 is 75 questions, 300 marks, 3 hours, computer-based.

Section A is 20 multiple-choice questions per subject, marked +4 / −1.

From 2025, Section B is 5 compulsory numerical questions per subject, marked +4 / −1.

Section B answers are typed on screen and rounded to the nearest integer.

JEE Main Paper 1 (B.E./B.Tech.) has 54 NTA units: 14 Mathematics, 20 Physics, 20 Chemistry.

NTA removed chapters from Physics, Chemistry and Mathematics in 2024; Chemistry lost the most.

January and April sessions; the better score is the one that counts.

Shift normalisation exists and is opaque.

The exam has 60 scored questions plus up to five unscored questions.

The sitting is 105 minutes, with no scheduled breaks.

The cut score is 73%. Salesforce publishes no per-section minimum.

Results are available immediately after the sitting.

Pass marks for Salesforce exams other than Agentforce Specialist.

The published weights are AI Agents 35%, Prompt Engineering 20%, Data Cloud for Agentforce 20%, Deployment Lifecycle 20%, Multi-Agent Interoperability 5%.

There are no prerequisites. Salesforce recommends six months of hands-on experience and the Platform Administrator and Platform App Builder credentials.

The exam is delivered online or at a test centre.

Salesforce requires a release maintenance module once a year. The certification expires if the deadline is missed.

A retake is $100.

The first-attempt registration fee.

The exam is published under the code AI-201.

Administrator became Platform Administrator, Advanced Administrator became Platform Administrator II, and Associate became Platform Foundations.

Pearson VUE for this exam includes a Flag for review control.

Most questions give you four answers that would all technically run; they differ by fit rather than by whether they work.

Readiness for this exam is shown as a percent band such as 68–76%.

Salesforce retired AI Associate on 2 February 2026.

Recertification is required every two years.

A renewal exam exists for at least some Databricks certifications.

A renewal exam exists for Data Engineer Associate: 25 scored questions, 60 minutes, USD 200, online only, requiring the current certification.

There is no renewal exam for the Data Engineer Professional. Recertifying means the full current exam.

There is no renewal exam for the Machine Learning Associate. Recertifying means the full current exam.

There is no renewal exam for the Machine Learning Professional. Recertifying means the full current exam.

There is no renewal exam for the Data Analyst Associate. Recertifying means the full current exam.

There is no renewal exam for the Generative AI Engineer Associate. Recertifying means the full current exam.

There is no renewal exam for the Associate Developer for Apache Spark. Recertifying means the full current exam.

Unscored items are included, are not identified, and extra time is factored into the limit.

Databricks does not publish a passing score for its certification exams, and revises passing scores as exams are updated.

The 80% pass mark widely quoted for Databricks certifications is the pass mark for Databricks badge assessments, which are free, unproctored and a separate product.

Readiness for Databricks exams is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

A 70% pass mark is widely quoted.

Readiness for the Data Engineer Professional is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Readiness for the Machine Learning Associate is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Readiness for the Machine Learning Professional is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Readiness for the Data Analyst Associate is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Readiness for the Generative AI Engineer Associate is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Candidates receive a domain-by-domain breakdown with their result.

Readiness for the Associate Developer for Apache Spark is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Databricks.

Code appears in SQL wherever SQL can express the concept, and in Python otherwise. Scala is not required.

45 questions / 90 minutes for Associate, 60 / 120 for Professional.

Data Engineer Associate: 45 scored questions, 90 minutes, multiple choice.

Data Engineer Associate questions are multiple choice.

Delivered online proctored or at a test centre.

No test aids.

Offered in English, Japanese, Brazilian Portuguese and Korean.

Data Engineer Professional: 59 scored questions, 120 minutes, multiple choice.

The current form has 59 scored questions in 120 minutes. The retired March 2025 guide said 60.

Data Engineer Professional questions are multiple choice.

Delivered online proctored or at a test centre.

No test aids are permitted, and the exam guide names API documentation specifically.

Offered in English, Japanese, Brazilian Portuguese and Korean.

Code examples on this exam are primarily in Python and SQL. Scala is not required.

Machine Learning Associate: 48 scored questions, 90 minutes, multiple choice.

Machine Learning Associate questions are multiple choice.

Delivered online proctored or at a test centre.

No test aids.

Offered in English, Japanese, Brazilian Portuguese and Korean.

All machine learning code on this exam is in Python. Data manipulation not specific to machine learning tasks may appear in SQL.

Machine Learning Professional: 59 scored questions, 120 minutes, multiple choice.

Machine Learning Professional questions are multiple choice.

Delivered online proctored or at a test centre.

No test aids are permitted.

The Machine Learning Professional is offered in English only.

Data Analyst Associate: 45 scored questions, 90 minutes, multiple choice.

Data Analyst Associate questions are multiple choice.

Delivered online proctored or at a test centre.

No test aids.

The Data Analyst Associate is offered in English only.

SQL throughout. This is the one Databricks exam in the catalogue with no Python on it.

Generative AI Engineer Associate: 45 scored questions, 90 minutes.

The exam includes multiple-selection items as well as multiple-choice, per the exam guide.

Delivered online proctored or at a test centre.

No test aids.

Offered in English, Japanese, Brazilian Portuguese and Korean.

Associate Developer for Apache Spark: 45 scored questions, 90 minutes, multiple choice.

Associate Developer for Apache Spark questions are multiple choice.

No test aids.

The certification page describes candidates who pass as being able to complete basic Spark DataFrame tasks using Python.

The Scala option has been withdrawn from this exam.

No formal prerequisite; Databricks recommends hands-on experience with the tasks in the guide.

Databricks recommends six months of hands-on work with the platform.

Databricks recommends one year of hands-on data engineering work on the platform. There is no prerequisite certification.

Databricks recommends six months of hands-on experience performing the machine learning tasks outlined in the exam guide. There is no prerequisite certification.

Databricks recommends one year of hands-on experience performing the machine learning tasks outlined in the exam guide. There is no prerequisite certification.

Databricks recommends six months as a data analyst. There is no prerequisite certification.

Databricks recommends six months of hands-on work with generative AI on Databricks. There is no prerequisite certification.

USD 200 per exam plus applicable local tax, full fee on retake. No retake voucher.

USD 200 plus applicable local tax.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

14 days between attempts, unlimited attempts, full fee each time. No retake voucher.

The credential is Databricks Certified Data Engineer Associate, at Associate level. Databricks assigns no public exam code.

The credential is Databricks Certified Data Engineer Professional, at Professional level. Databricks assigns no public exam code.

The credential is Databricks Certified Machine Learning Associate, at Associate level. Databricks assigns no public exam code.

The credential is Databricks Certified Machine Learning Professional, at Professional level. Databricks assigns no public exam code.

The credential is Databricks Certified Data Analyst Associate, at Associate level. Databricks assigns no public exam code.

The credential is Databricks Certified Generative AI Engineer Associate, at Associate level. Databricks assigns no public exam code.

The credential is Databricks Certified Associate Developer for Apache Spark, at Associate level. Databricks assigns no public exam code.

Domain names and weights for any Databricks exam.

The Data Engineer Associate has seven domains weighted 6 / 21 / 22 / 16 / 10 / 10 / 15.

The Data Engineer Professional has ten domains weighted 22 / 7 / 10 / 5 / 10 / 13 / 10 / 7 / 10 / 6.

The Machine Learning Associate has four domains weighted 38 / 19 / 31 / 12.

Databricks Machine Learning, the platform domain, is 38% of the exam — larger than Model Development at 31%.

A four-domain split of Databricks ML 29% / ML Workflows 29% / Spark ML 33% / Scaling ML Models 9% describes a retired version of this exam.

A six-domain split (Databricks Tooling 20 / Data Processing 30 / Data Modeling 20 / Security and Governance 10 / Monitoring and Logging 10 / Testing and Deployment 10) describes a retired version of this exam.

The Machine Learning Professional has three domains: Model Development 44%, ML Ops 44%, Model Deployment 12%.

Model Development and ML Ops are weighted identically at 44% each, 88% of the exam between them.

A four-domain split of Experimentation / Model Lifecycle Management / Model Deployment / Solution and Data Monitoring describes a retired version of this exam.

The Data Analyst Associate has nine domains weighted 11 / 8 / 5 / 20 / 15 / 16 / 12 / 5 / 8.

Dashboards (16%) and AI/BI Genie spaces (12%) are 28% of the exam between them, against 20% for executing queries.

A five-domain split opening with Databricks SQL at 22% describes a retired version of this exam.

The Generative AI Engineer Associate has six domains weighted 14 / 14 / 30 / 22 / 8 / 12.

The Associate Developer for Apache Spark has seven domains weighted 20 / 20 / 30 / 10 / 10 / 5 / 5.

The DataFrame/DataSet API domain is 30% of the exam; architecture and Spark SQL are 20% each.

The 4 May 2026 version renames platform features on the form: Lakeflow Jobs (formerly Workflows), Lakeflow Spark Declarative Pipelines (formerly Delta Live Tables), Git Folders (formerly Databricks Repos), Declarative Automation Bundles (formerly Databricks Asset Bundles).

The current Data Engineer Professional outline has been live since 3 July 2026.

Databricks' localized certification pages can carry retired blueprints. On 18 September 2026 the /jp/ Data Analyst page still listed the five-domain split opening with Databricks SQL at 22%.

The current version of this exam went live on 18 March 2026.

Guides were revised in July 2025, September 2025 and May 2026.

Databricks does not offer practice exams and states that practice-exam results are often misleading.

Associate Cloud Engineer: 50–60 multiple choice and multiple select questions, 2 hours.

Cloud Digital Leader: 50–60 multiple choice and multiple select questions, 90 minutes.

Online-proctored or onsite-proctored at a testing centre.

Offered in English, Japanese, Spanish and Portuguese.

The credential is Google Cloud Certified — Associate Cloud Engineer, at Associate level.

The current English Associate Cloud Engineer standard exam guide went live on 30 June 2026 and includes Gemini Cloud Assist, Gemini CLI, Google Antigravity and Application Design Center.

Google does not publish a pass mark for Associate Cloud Engineer. The result is pass or fail.

Google does not publish a pass mark for Cloud Digital Leader. The result is pass or fail.

Readiness for Associate Cloud Engineer is measured against an assumed threshold of 80% weighted accuracy, set by Acedify, not by Google.

Associate Cloud Engineer sections are approximately 20% setup, 30% plan and implement, 30% operate, 20% access and security.

USD 125 plus tax where applicable.

No prerequisites.

Google recommends 6+ months hands-on experience with Google Cloud.

Associate Cloud Engineer is valid for 3 years.

Associate Cloud Engineer can be renewed via the standard exam, a 20-question one-hour renewal exam for USD 75, or designated Google Skills courses that extend validity by one year. Foundational and Associate renewal may start 180 days before expiry.

Cloud Digital Leader renewal exam: 20 multiple choice and multiple select questions, 45 minutes, USD 60. Skills-course renewal extends validity by one year.

Google publishes a retake and attempt-cap policy for these exams.

Writing Task 2 is widely held to carry about twice the weight of Task 1 when the two tasks are combined into one Writing band.

A full Speaking band on Acedify combines Part 1, Part 2 and Part 3 by estimated candidate-speech share (Part 1 and Part 3 carry more speech than Part 2). Examiners score the interview as a whole; this weighting is an estimate, not an official rule, and it is not equal thirds.

Displayed IELTS skill and overall ranges are Acedify estimates from a model whose constants are provisional (status SEED).

The Writing range combines Task 1 and Task 2 at 1:2 (see ielts-writing-task-weighting).

Speaking ranges built from fewer than three parts are labelled with the parts covered ("Part 2 of 3", "Parts 1 and 2 of 3").