This article summarises the Claude Certified Architect Professional (CCAR-P) exam guide published by Anthropic: what the certification assesses, the exam format, what the domain weights mean, the objectives under each domain, and how CCAR-P differs from Claude Certified Architect Foundations (CCAR-F).
What CCAR-P assesses
Anthropic's exam guide states that the purpose of the certification is "to provide an independent assessment of the knowledge, skills, and abilities required to architect Claude-based solutions competently in production environments."
The exam is intended for mid- to senior-level technical professionals such as solution architects, AI/ML engineers, technical leads and senior software engineers who design and deliver production AI solutions with large language models. The guide recommends:
- A foundation in software engineering best practices
- 3+ years of experience in systems architecture or platform engineering
- 6+ months of hands-on experience with Claude or comparable LLM-based systems in production
- Experience delivering end-to-end systems from discovery through deployment
There are no mandatory prerequisites or required courses. The experience above is recommended, not required.
Exam format
| Item | Detail |
|---|---|
| Exam code | CCAR-P |
| Questions | 63 |
| Question types | Multiple choice and multiple response. Each question states how many answers to select. |
| Time | 120 minutes |
| Delivery | Proctored, online or at a test centre |
| Passing score | Scaled score of 720 on a scale of 100 to 1,000 |
| Result | Pass or fail, scaled score, and percent correct by domain |
| Exam fee | US$175 |
| Validity | 12 months from the award date |
A scaled score of 720 does not mean 72% of questions answered correctly.
What domain weights mean
The exam guide divides the exam content into seven domains and gives each domain a weight. The guide explains the weights as follows:
"Weights reflect the relative importance of each domain to competent performance as determined through the job task analysis. The percentages indicate the approximate proportion of scored items drawn from each domain."
In practice, a domain weight tells you roughly what share of the scored questions come from that domain. Multiplying the weight by 63 gives an estimate of the number of questions. The actual number on your exam can differ.
| Domain | Domain weight | Estimated questions |
|---|---|---|
| Integration | 19% | 12 |
| Solution Design & Architecture | 17% | 11 |
| Evaluation, Testing & Optimization | 16% | 10 |
| Governance, Safety & Risk Management | 14% | 9 |
| Stakeholder Communication & Lifecycle Management | 14% | 9 |
| Claude Models, Prompting & Context Engineering | 13% | 8 |
| Developer Productivity & Operational Enablement | 7% | 4 |
No single domain carries more than 19%. Six of the seven domains sit between 13% and 19%.
Your result shows percent correct for each domain, so you can see which domains to study again if you need to retake the exam.
Objectives in each domain
The guide lists objectives under each domain. The objectives do not have separate weights. The lists below follow the guide's wording.
Solution Design & Architecture (17%)
- Translate business problems into Claude-based AI solutions
- Design end-to-end architectures (input, processing, output, feedback loops)
- Select architectural patterns (workflow, agentic, augmented LLM)
- Design multi-agent systems and orchestration strategies
- Apply decomposition techniques for complex problem solving
- Align solutions to business value pillars (efficiency, transformation, productivity, cost, performance SLAs)
Claude Models, Prompting & Context Engineering (13%)
- Select Claude models based on trade-offs
- Design system prompts, templates and guardrails
- Apply prompt engineering techniques (zero-shot, few-shot, chain-of-thought)
- Optimize context windows and manage token usage
- Implement prompt reuse strategies (caching, modular prompts, Skills)
Integration (19%)
- Evaluate tool and agent configuration for capability bloat
- Analyze authentication and authorization requirements to identify security gaps
- Evaluate accuracy-latency trade-offs and justify configuration decisions
- Analyze observability challenges and select monitoring strategies at scale
- Design a RAG pipeline with appropriate chunking and indexing strategies
- Apply retrieval strategies matched to data shape and query pattern
- Evaluate connection protocols and select an integration mechanism (MCP, API/CLI, agent-to-agent)
- Evaluate progressive discovery vs. monolithic context strategy
Evaluation, Testing & Optimization (16%)
- Define evaluation metrics (accuracy, latency, cost, safety, security)
- Design evaluation datasets and test frameworks using mixed methodologies
- Conduct A/B testing and iterative improvements
- Diagnose system issues (prompt failure, hallucinations, model mismatch)
- Optimize token usage, latency and cost-performance trade-offs
- Monitor system performance using logging and observability tools
Governance, Safety & Risk Management (14%)
- Implement guardrails and safety controls
- Identify risks, limitations and failure modes of LLM systems
- Apply human-in-the-loop validation strategies
- Ensure compliance with regulations (for example GDPR, HIPAA, FedRAMP)
- Address ethical AI considerations (bias, fairness, transparency)
Stakeholder Communication & Lifecycle Management (14%)
- Conduct structured discovery and requirement gathering
- Communicate architectural decisions and trade-offs
- Manage stakeholder feedback loops and expectation alignment, including SLAs
- Document architectures and provide implementation guidance
- Support lifecycle phases (discovery, design, handoff, monitoring, iteration)
Developer Productivity & Operational Enablement (7%)
- Configure Claude tools and environments for teams (for example Claude Code)
- Improve developer workflows using AI-assisted tooling
- Support debugging and operational issue resolution
How CCAR-P differs from CCAR-F
Both exams are for architects. The two exam guides differ in these ways:
| Item | CCAR-F | CCAR-P |
|---|---|---|
| Questions | 60 | 63 |
| Exam fee | US$125 | US$175 |
| Recommended experience | 6+ months building with Claude APIs, Agent SDK, Claude Code and MCP | 3+ years in systems architecture or platform engineering, and 6+ months with Claude or comparable LLM systems in production |
| Structure described in the guide | 4 scenarios picked from a bank of 6 | No scenario structure described |
| Domains | 5 | 7 |
| Largest domain | Agentic Architecture & Orchestration (27%) | Integration (19%) |
Three CCAR-P domains have no CCAR-F equivalent: Evaluation, Testing & Optimization; Governance, Safety & Risk Management; and Stakeholder Communication & Lifecycle Management.
CCAR-F task statements name specific configuration: stop_reason values, tool_choice settings, .mcp.json, CLAUDE.md levels, CLI flags. CCAR-P objectives are about design decisions and their trade-offs: RAG pipelines, authentication gaps, evaluation metrics, regulatory compliance and stakeholder communication.
Example question
This is an original Timo practice question on the Evaluation, Testing & Optimization domain. It is not an official exam question.
A team has built a contract review assistant that will replace a manual first review. The client requires at least 95% accuracy on each of its five contract types, measured before go-live. The team tested the assistant on 1,000 contracts that lawyers had already reviewed. Overall accuracy was 96%.
What should the team do next?
A. Go live, because overall accuracy of 96% is above the 95% requirement.
B. Measure accuracy separately for each contract type on the reviewed set, and fix any type below 95% before go-live.
C. Switch to the most capable Claude model to raise accuracy further, then go live.
D. Go live, and track complaints from the client's lawyers to find weak contract types.
Answer: B. The requirement is 95% for each contract type. An overall figure of 96% can hide one type that is well below 95%, so the team must measure each type separately.
- A uses the overall figure, which does not show accuracy for each contract type.
- C changes the model without measuring accuracy for each contract type, so the requirement is still not checked.
- D measures after go-live, but the client requires the measurement before go-live.
More questions in this format are in the free 20-question CCAR-P practice test, which does not require sign-up. For the exam format and registration steps, see the CCAR-P exam guide on Timo Labs.
Top comments (0)