pathfinder ES
All certifications
Anthropic CCAR-F badge

Anthropic · CCAR-FAssociateAI engineeringRestricted

Claude Architect Foundations

Agent architecture; Tools / MCP; Claude Code workflows; Prompts / output structure; Context / reliability

40 original practice questions, in English and Spanish. They are not real exam questions.

Exam facts

PriceUS$125Exam only
Exam length2 h
Questions60
Passing score720 scaled (100–1,000)
Valid for12 months
FormatQuestion-based exam
RegistrationRestricted eligibility

Exam domains

Agentic Architecture & Orchestration: 27%Tool Design & MCP Integration: 18%Claude Code Configuration & Workflows: 20%Prompt Engineering & Structured Output: 20%Context Management & Reliability: 15% 5 domains
  • Agentic Architecture & Orchestration27%
  • Tool Design & MCP Integration18%
  • Claude Code Configuration & Workflows20%
  • Prompt Engineering & Structured Output20%
  • Context Management & Reliability15%

Weights from the official exam guide v1.0 (effective July 2026), linked from the certification page.

Good to know

  • Restricted to Claude Partner Network organizations: registration needs a company email domain recognized on the partner record. A personal Claude subscription is not enough.
  • 60 multiple-choice and multiple-response items in 120 minutes, proctored online or at a Pearson VUE test center. 12-month validity; on-time renewal is a free, non-proctored assessment. Each sitting draws 4 scenarios from a bank of 6.
  • AI during the exam: The reviewed Claude certification rules prohibit AI assistance during the closed-book exam.
  • Retakes need a wait of 14, 30 and then 90 days after each failed attempt, with at most four attempts in 12 months; each attempt pays the fee.

Sources

Checked 3 Nov 2026 against the issuer. Prices change and vary by country. Confirm with the issuer before you pay.

Badge image: the issuer's official badge

7 candidate reports

Self-reported. Useful for how people prepared, not for pass rates.

Udacity: a practitioner’s six-week account · Blog

Describes scenario judgment and limitations of generated practice.

Red Airship: an eight-person team account · Blog

Employer reports eight passes and recommends practical work.

Roman Gorge: daily experience still left gaps · Blog

Reports a pass with stronger orchestration skills but weaker hooks/enforcement knowledge.

freeCodeCamp / Andrew Brown: Claude course · Blog

Hands-on video resource covering agent and API-related workflows.

Early Architect experience: production judgment · Reddit

Emphasizes architectural reasoning and explaining why plausible alternatives fail, rather than memorizing feature definitions.

Second-attempt pass after focused repair · Reddit

Claims an initial 696 and later 809. Shifted from collecting resources to weak-domain practice and defending decisions against distractors.

Two weeks can still mean sixty hours · Reddit

Reports a pass after intensive scenario practice and method changes. Shows why a two-week headline cannot establish a low study-hour requirement.

Similar certifications