Written by the AgileFever editorial team. Facts checked on November 5, 2026 against the official pages of the organizations named in this post. AgileFever sells training for some of the topics covered here, and we say so wherever it applies. Every exam figure below comes from the exam owner’s own page. Hours of study are our estimates and are labelled that way.
The CCA-F certification and the cloud AI certifications from AWS, Microsoft and Google Cloud are not substitutes. They measure different things, cost different amounts, and open to different people. This post puts the published facts side by side, then maps them to five roles so you can pick a first exam and plan the second.
CCA-F certification vs cloud AI certifications, side by side
We picked the currently active AI exams at two levels from each vendor: a foundational exam and a role-based one that covers generative AI and agents. Where a vendor page does not print a number, the cell says so.
| Exam | Questions | Time | Fee (USD) | Validity | Prerequisites |
|---|---|---|---|---|---|
| Anthropic, Claude Certified Architect – Foundations (CCA-F) | 60 | 120 min | $125 (50% off for Select, Preferred and Global Premier partners) | 12 months, free on-time renewal | Must work at a Claude Partner Network organization; company email; age 18+ |
| AWS Certified AI Practitioner | 65 | 90 min | $100 | 3 years | None. AWS recommends its cloud essentials course for newcomers |
| AWS Certified Generative AI Developer – Professional | 75 | 180 min | $300 | 3 years | None required. AWS expects 2+ years building on AWS and 1 year with generative AI |
| Microsoft Azure AI Fundamentals (AI-901) | Not printed | Not printed | Priced by country | Does not expire | Python syntax and Azure resource familiarity expected. Pass mark 700 |
| Microsoft Azure AI Apps and Agents Developer Associate (AI-103) | Not printed | 120 min | Priced by country | Renewed yearly, free, open-book | Python app development experience expected |
| Google Cloud Generative AI Leader | 50 to 60 | 90 min | $99 | 3 years | None. Any job role |
| Google Cloud Professional Machine Learning Engineer | 50 to 60 | 120 min | $200 | Renewal adds 2 years | None required. Google recommends 3+ years in industry, 1+ on Google Cloud |
Sources: Anthropic Partner Academy certifications FAQ and exam guide, AWS certification and recertification pages, Microsoft Learn exam, certification and renewal pages, Google Cloud certification pages and renewal FAQ, all read on November 5, 2026. Vendors can change fees and formats. Confirm on the exam page before you pay. Cloud exam fees exclude local tax and currency fees.
Two details hide in the table. First, the CCA-F certification is the only exam here with an eligibility gate. Anthropic’s FAQ says certification is currently available only to organizations in the Claude Partner Network. Second, the validity periods differ by a factor of three: you renew CCA-F every 12 months, the AWS and Google foundational exams every three years. Read the Anthropic Partner Academy certification page before you plan around either number.
What each exam measures
The exam title tells you less than the domain list does. Here is what each owner says the exam tests.
Source: Anthropic CCA-F exam guide v1.0, checked November 5, 2026. Weights are approximate.
- Agentic Architecture & Orchestration: 27
- Claude Code Configuration & Workflows: 20
- Prompt Engineering & Structured Output: 20
- Tool Design & MCP Integration: 18
- Context Management & Reliability: 15
The CCA-F certification is scenario-based. Four scenarios are drawn from a bank of six, and every question asks you to choose a design for a described situation. The guide excludes fine-tuning, vision, computer use, streaming and cloud-provider configuration. The target candidate has six or more months of building with the Claude API, Agent SDK, Claude Code and MCP.
- AWS AI Practitioner: foundational AI and generative AI concepts and AWS services. AWS lists no prerequisites.
- AWS Generative AI Developer – Professional: production generative AI applications on AWS, for people with the experience listed in the table.
- Azure AI-901: Microsoft splits it into identifying AI concepts and capabilities (40 to 45%) and implementing AI solutions with Microsoft Foundry (55 to 60%).
- Azure AI-103: planning and managing an Azure AI solution, generative AI and agentic solutions, computer vision, text analysis and information extraction.
- Google Generative AI Leader: business-level knowledge: gen AI fundamentals, Google Cloud’s offerings, techniques to improve model output and business strategy. No hands-on experience is needed.
- Google Professional ML Engineer: building, productionizing and monitoring ML and generative AI solutions on Google Cloud, including Vertex AI Agent Builder.
The pattern is clear. The cloud exams test a platform: its services, security, deployment and cost. CCA-F tests a way of building agents with one model family, and it deliberately leaves the platform out. A person who knows both can design the agent and run it on a cloud. That is why the exams pair well, not compete.
Exam fees and time in the seat
Fee and length are the two numbers every vendor publishes in dollars and minutes, so we can compare them directly. The Azure exams are left out of the fee chart because Microsoft prices them by country and prints no single figure.
Sources: Anthropic Partner Academy FAQ, AWS, Google Cloud exam pages, read November 5, 2026. Partner discounts and local taxes are not included.
Sources: the same owner pages, plus the Microsoft Learn AI-103 page. Seat time with check-in is longer.
The CCA-F certification sits in the middle on both axes. It costs more than the foundational cloud exams and less than the professional ones. The fee is not what separates these choices. Eligibility and fit are.
Which certification first, by role
Use this table as a starting point. It reflects the published audiences and rules above. The reasoning in the last column is our judgment, not vendor guidance, and nobody publishes pass rates or hiring data that would settle it.
| Role | Take first | Take second | Reason |
|---|---|---|---|
| Solution architect at a Claude partner firm | CCA-F | A cloud role-based exam for the platform your clients use | CCA-F matches the architect scenarios you sell, and your firm’s partner standing counts certified people. |
| Solution architect elsewhere | The cloud AI exam for your main cloud | CCA-F once your employer joins the network | You cannot register for CCA-F yet. A cloud exam is open to you now. |
| Developer building agents | CCA-F if eligible, else the cloud developer exam (AWS Generative AI Developer, Azure AI-103) | The other one | Both test building. CCA-F tests design choices on Claude, the cloud exams test the platform. |
| DevOps or platform engineer | The cloud AI exam on your platform | CCA-F if you also run Claude Code in CI/CD | Deployment, identity and cost sit with the cloud exams. CCA-F covers Claude Code in automation pipelines only. |
| Data or ML engineer | Google Professional ML Engineer or AWS AI Practitioner, then your platform’s developer exam | CCA-F for agent design | The ML-specific exams cover model pipelines. CCA-F skips fine-tuning and training. |
| Consultant or pre-sales | Foundational: Google Generative AI Leader or AWS AI Practitioner | CCA-F at a partner firm | Both foundational exams list no prerequisites and suit non-coders. CCA-F then adds client-facing credibility where it counts. |
Yes: CCA-F is open to you. No: a cloud exam is your first step.
Yes: aim at a developer-level exam. No: a foundational exam such as AI Practitioner, AI-901 or Generative AI Leader fits.
Pick that cloud’s exam. Learning a platform you will not use adds cost and no signal.
Yes: add CCA-F as the second or first exam. Its 12-month validity suits people who build now.
No: build first. Both CCA-F and the professional cloud exams assume it.
How to combine them over a year
Here is one sequence for someone at a partner firm who works with a single main cloud. The hours are our estimates for a working IT professional who already builds with the tools. The vendors publish no preparation hours, so treat the numbers as planning figures, not promises.
| Months | Exam | Our study-hours estimate | Why this slot |
|---|---|---|---|
| 1 to 2 | Foundational cloud AI exam (AI Practitioner, AI-901 or Generative AI Leader) | 15 to 25 hours | Low fee, long validity, builds the vocabulary of your platform. |
| 3 to 5 | CCA-F | 30 to 40 hours | Our own CCA-F study plan runs 30 to 37 hours. Add time if you have not used Claude Code or MCP. |
| 6 to 10 | Role-based cloud exam (Generative AI Developer, AI-103 or Professional ML Engineer) | 60 to 100 hours | Longest exams, widest experience expectation. Do it after you have shipped something. |
| 11 to 12 | CCA-F renewal (free, non-proctored) | 5 to 10 hours | Renew inside the 12-month window. A lapse means the full exam at the full fee. |
That adds up to roughly 110 to 175 hours across the year, and about $525 in published exam fees at list price if you take AI Practitioner, CCA-F and the AWS Generative AI Developer exam ($100 + $125 + $300). Partner discounts lower the CCA-F line.
Skip the foundational exam if you already hold an associate cloud certification and build daily. It adds a badge, not new knowledge.
Where AgileFever fits
AgileFever sells live training in this area, and none of it is a substitute for the exam owners’ own study guides. The prep program for the CCA-F certification is 16 hours of live instruction built around the five exam domains. As listed on November 5, 2026, it costs $450 or Rs 25,000 (before 18% GST for buyers in India). See the Claude Certified Architect Foundations program.
The Enterprise AI Agents programs for each cloud are build courses: 12 hours of live training, 6 modules and 10 hands-on labs. They teach you to build, deploy and operate agents on the platform. They are not written as vendor exam prep, but the hands-on practice overlaps with what the role-based exams expect. As listed on November 5, 2026, each costs $199 or Rs 16,000 before GST. Pick your platform:
- Enterprise AI Agents on AWS, next batch Nov 23 to Nov 26, 2026.
- Enterprise AI Agents on Azure, next batch Nov 16 to Nov 19, 2026.
- Enterprise AI Agents on Google Cloud, next batch Nov 30 to Dec 3, 2026.
Additional discounts are available. Contact us to know more.
If you cannot yet decide which cloud, our Azure vs AWS vs Google Cloud comparison for AI agents and guide for cloud engineers help. For CCA-F itself, start with the cost, eligibility and exam details guide, the syllabus walk-through and practice questions without dumps.
Who should skip the CCA-F certification for now
- Anyone whose employer is not in the Claude Partner Network. You cannot register.
- Anyone with no hands-on Claude work. The exam guide describes a candidate with six or more months of building.
- Anyone who needs a credential recognised across all clouds. The CCA-F certification is new, narrow and Claude-specific, and no hiring data exists for it yet.
- Anyone who wants a general AI credential. A foundational cloud exam covers broader concepts.
Questions people ask
Should I take the CCA-F certification or a cloud AI certification first?
If your employer is in the Claude Partner Network and you build with Claude, take CCA-F first. Otherwise take the cloud AI exam for the platform you use, because CCA-F registration is limited to partner organizations. The two answer different questions, so many people eventually hold both.
Is the CCA-F certification harder than AWS AI Practitioner?
They target different people. AI Practitioner is foundational, with 65 questions in 90 minutes and no prerequisites. CCA-F has 60 scenario-based questions in 120 minutes and assumes six months of building with Claude. No vendor publishes pass rates, so any difficulty ranking is opinion.
Do I need a cloud certification before the CCA-F certification?
No. Anthropic lists no cloud prerequisite, and the exam guide excludes cloud-provider configuration. The requirement is eligibility through a partner organization and hands-on Claude experience.
How long does each certification stay valid?
The CCA-F certification is valid 12 months, with a free non-proctored renewal. AWS AI Practitioner, AWS Generative AI Developer and Google Generative AI Leader are valid 3 years. Microsoft associate certifications renew yearly for free, and Azure AI Fundamentals does not expire. Google Professional exams add 2 years on renewal.
Can an independent consultant take the CCA-F certification?
Not today. Anthropic’s FAQ says certification is currently available only to organizations in the Claude Partner Network, and registration needs a company email on a recognised partner domain. Independent consultants can take the Claude Academy courses or a cloud exam instead.
What does the Azure exam cost?
Microsoft’s pages price Azure exams by the country where you take them and print no single figure. Check the exam page for your country before you plan the budget.
