This is Anthropic's first technical certification for people who design and ship production-grade agentic applications with Claude. This guide covers the official exam format, tested domains, and a study plan, based on Anthropic's published Exam Guide.
Last updated July 2026 · figures verified against the official Exam Guide.
Exam quick facts
| Field | Value |
|---|---|
| Official name | Claude Certified Architect — Foundations (CCA-F) |
| Format | 60 multiple-choice questions (1 correct + 3 incorrect options), across 4 scenarios drawn from a bank of 6 |
| Duration | 120 minutes, single sitting, no external resources |
| Passing score | 720 on a 100–1,000 scaled score |
| Exam fee | $125 per attempt |
| Credential validity | 12 months from the date it's awarded |
| Delivery | Online proctored or at a test center, via Anthropic Partner Academy (Skilljar) |
These figures were checked directly against Anthropic's official "Exam Guide," published on the Partner Academy (Skilljar) — not third-party blogs. Since the certification is new (launched March 2026) and pricing may carry special terms for Claude Partner Network members, always confirm current pricing on the official page before purchasing your attempt.
What is the CCA-F?
Claude Certified Architect — Foundations is Anthropic's entry-level technical certification, built to validate that a professional can design and deploy agentic applications with Claude at production scale — not just use the chat interface, but architect agent systems, tools, and context.
Per the official Exam Guide, the target candidate is a solution architect with hands-on experience building with the Claude Agent SDK, Claude Code, MCP, and the Claude API — typically 6+ months of practical experience.
It launched inside the Claude Partner Network, Anthropic's partner program, but the exam itself is individual: the credential belongs to the person who passed, not their employer.
Domains covered on the exam
Per the Content Outline in the official Exam Guide, CCA-F is organized into five competency domains, weighted by Anthropic itself. Below is each domain with the task statements that come up most — plus how AICert Study's own question bank (1,049 questions) is organized across the same five domains.
| Domain | Weight |
|---|---|
| Agentic Architecture & Orchestration | 27% |
| Claude Code Configuration & Workflows | 20% |
| Prompt Engineering & Structured Output | 20% |
| Tool Design & MCP Integration | 18% |
| Context Management & Reliability | 15% |
Agentic Architecture & Orchestration (27%)
The agentic loop lifecycle and stop_reason handling, hub-and-spoke coordinator/subagent orchestration, context passing and spawning between agents, Agent SDK hooks for tool-call interception, task decomposition, and session state, resumption, and forking.
Claude Code Configuration & Workflows (20%)
CLAUDE.md hierarchy and modular organization (@import, .claude/rules/), custom slash commands and skills, path-specific rules, choosing plan mode vs. direct execution, iterative refinement techniques, and integrating Claude Code into CI/CD pipelines.
Prompt Engineering & Structured Output (20%)
Designing prompts with explicit criteria to reduce false positives, few-shot prompting for consistency, enforcing structured output via tool use and JSON schemas, validation/retry feedback loops, batch processing with the Message Batches API, and multi-instance review architectures.
Tool Design & MCP Integration (18%)
Designing clear, non-overlapping tool interfaces, structured error responses for MCP tools, distributing tools across agents with tool_choice, integrating MCP servers into Claude Code, and choosing the right built-in tools (Read, Write, Edit, Bash, Grep, Glob).
Context Management & Reliability (15%)
Managing conversation context across long interactions without losing critical details (the "lost in the middle" effect), plus the broader reliability and escalation practices covered under this domain.
The weights above are exactly what's published in Anthropic's official Exam Guide (Content Outline). Our 1,049-question bank uses the same five domains, so you can practice in the same proportions as the real exam.
How to study: a 4-step plan
1. Take the free mini practice exam (no signup)
Try 10 real exam-style questions right on the homepage, no account required — the fastest way to gauge the difficulty level.
2. Create an account to unlock the full 60-question exam
After the mini exam, sign up to see correct answers, detailed explanations, and unlock the full free practice exam — timed, with results shown at the end.
3. Switch to Study mode, filtered by domain
Use dynamic practice exams filterable by domain (e.g. just Tool Design & MCP Integration, or just Agentic Architecture) with instant feedback and explanations linked to official Anthropic documentation.
4. Repeat the domains you get wrong most
The per-domain performance report shows exactly where to focus before booking the official exam, instead of generically restudying what you already know.
Try the free CCA-F practice exam →
Frequently asked questions
How much does the official Anthropic exam cost?
Per the official Exam Guide and registration page on Anthropic Partner Academy (Skilljar), the fee is $125 per attempt. Confirm current pricing on the official page before purchasing, since terms may differ for Claude Partner Network members.
How long is the exam?
120 minutes, single sitting, to answer 60 multiple-choice questions organized across 4 scenarios drawn from a bank of 6.
How long is the credential valid?
12 months from the date it's awarded, per the official Exam Guide.
Is AICert Study's practice exam official Anthropic content?
No. AICert Study is an independent platform, not affiliated with Anthropic. Our questions are AI-assisted, built from Anthropic's public documentation to help with preparation — they don't replace the official exam or guarantee a passing score.
Is the certification worth it?
If you professionally design or deploy agentic applications with Claude — or want to prove that skill to clients or employers — it's a differentiator that's still uncontested, given how recently it launched.
About this guide
AICert Study is an independent product with no affiliation, sponsorship, or connection to Anthropic. The quick facts and domain data were checked against Anthropic's official Exam Guide; the rest of this guide was written with AI assistance and may contain inaccuracies — always confirm pricing, format, and rescheduling policy on Anthropic's official source before booking your exam.
See the official page and Exam Guide on Anthropic Partner Academy →
Ready to test your level? Start now, free →