Claude Certified Developer – Foundations: Frequently Asked Questions

It's an independent assessment of the knowledge, skills, and abilities needed to build Claude-based applications competently at a foundational level. Passing signals you can independently own or significantly contribute to building, integrating, and shipping Claude-powered systems — specifically: building agents with the Claude Agent SDK or custom agent loops, integrating Claude via the API and client SDKs (streaming, error handling, multi-format input), operating Claude Code (Skills, CLAUDE.md, settings.json, plugins), writing prompts and applying context engineering, running evals and debugging via trace analysis, selecting model tiers and managing cost/latency, applying security guardrails against prompt injection, and building custom tools and MCP servers.

53 scored multiple-choice and multiple-response items in 120 minutes. That's the fewest items of any of the four Claude certifications — Architect Foundations and Associate Foundations each use 60 items, and Architect Professional uses 63. With the same 120-minute window as the others, CCDV-F actually gives you slightly more time per item (roughly 2.3 minutes) than the higher-item-count exams.

$125 USD per attempt, charged again for any retake. It's the same fee as Architect Foundations, higher than Associate Foundations ($99), and lower than Architect Professional ($175).

A scaled score of 720 out of a 100–1,000 range — identical scoring scale and passing threshold across all four Claude certifications.

If you're inside or close to the target profile — roughly 1-5 years of software engineering experience plus 6+ months hands-on with Claude — the exam tests things you've likely already touched in production: API integration, tool use, MCP, prompt/context handling, and basic security guardrails. It's not designed as a trick-question exam; it's designed to confirm applied competence against a defined blueprint. The heavy skew toward Applications and Integration (33.1%) means solid, practical API/SDK experience matters more than memorized trivia.

There's no official prep-time figure from Anthropic, and this site won't invent one. A reasonable approach: if you're already shipping Claude integrations regularly, focus review time on the domains you touch least — for many engineers that's Security and Safety (8.1%), Eval/Testing/Debugging (2.6%), or Claude Code specifics (3.1%) like Skills authoring and settings.json configuration, since day-to-day work often skips those even while spending plenty of time on core API integration.

No mandatory prerequisites. Registration is open to anyone through the Anthropic Partner Academy. The 1-5 years SWE / 6-month Claude target profile describes who the content is calibrated for, not a gate on who can sit the exam.

Either is sufficient — the exam guide states candidates should be proficient in Python and/or TypeScript, the two languages with first-party Claude SDKs. You don't need both languages, but you do need comfort with SDK-level code (typed responses, async/streaming, structured tool schemas) in at least one.

You can retake it after a waiting period that lengthens with each failed attempt: 14 days after the first failure, 30 after the second, 90 after the third, with a maximum of four attempts in any rolling 12-month period. Each attempt costs the full $125 fee. Full detail is on the [shared policy reference](/claude-certifications/policies).

12 months from the date awarded. Renewing before expiration is a free, non-proctored assessment through the Anthropic Partner Academy. If it lapses, you need a full paid retake at $125 to regain certified status. See [validity](/claude-certified-developer/validity) for a worked timeline.

Anthropic's Partner Academy is the authoritative source for the exam guide and registration. Beyond that, a handful of independent resources exist in the wild, including at least one paid multi-certification exam-prep platform that offers a gated free sample test alongside paid full access. This site's content — covering the full domain blueprint, eligibility, cost, and FAQ — is free and complete, with no gate, as a straightforward alternative to piecing together fragments across paid platforms.

They're related but distinct in scope, not tiers of the same thing. Developer Foundations validates building, integrating, and shipping application-level Claude code (heaviest domain: Applications and Integration at 33.1%). [Architect Foundations](/claude-certified-architect) validates broader system design and orchestration (heaviest domain: Agentic Architecture & Orchestration at 27%). Neither is a prerequisite for the other, and it's reasonable to hold both if your role spans both scopes.

Associate Foundations targets non-technical or lightly technical professionals who use Claude as a productivity tool — prompting, evaluating output, workflow integration — without writing application code. CCDV-F explicitly excludes that audience and targets engineers who build and ship Claude-powered software. If your job doesn't involve writing or shipping code, [Associate Foundations](/claude-certified-associate) is the better fit.

A pass/fail result tied to your scaled score (100–1,000, with 720 as the pass threshold), plus a percent-correct breakdown by each of the eight domains. Because the domains are so unevenly weighted (33.1% down to 2.6%), a shortfall in a light domain affects your overall score much less than the same shortfall in Applications and Integration or Model Selection and Optimization. See [validity](/claude-certified-developer/validity) for a full walkthrough of interpreting this.

No, and it isn't designed to. It validates knowledge and applied skill against a fixed blueprint at a point in time — it doesn't validate judgment under your team's specific scale, codebase, or incident history. Treat it as a complement to a portfolio of shipped work, not a substitute for one. Anthropic's own framing supports this: the credential is about demonstrated competence, not a replacement for hands-on delivery.

Through the Anthropic Partner Academy, which handles exam registration and checkout, followed by Pearson VUE for scheduling your proctored session (online or at a test center). The full 6-step registration flow is documented on the [shared policy reference](/claude-certifications/policies).

Both options are available. The exam is delivered by Pearson VUE, and you choose between online proctoring or an in-person test center when scheduling after registration.

Applications and Integration, at 33.1% of the exam — more than double the next-heaviest domain (Model Selection and Optimization at 16.8%). Together those two domains represent just under half the entire exam, so strength there has an outsized effect on your scaled score compared to the four lightest domains (Security and Safety, Tools and MCPs, Claude Code, and Eval/Testing/Debugging) combined.

Yes — appeals must be submitted within 14 days of being notified of the decision, or within 14 days of the exam date for a concern about the result itself, through Pearson VUE support. Details are on the [shared policy reference](/claude-certifications/policies).

All 53 scored items are multiple-choice or multiple-response — there's no live-coding component, no submitted repository, and no lab environment. The questions are expected to be scenario-based given the domain descriptions (a described integration bug, an agent looping incorrectly, a context-window degradation issue), so real hands-on experience helps even though you're not writing code during the exam itself.

Spending roughly equal study time across all eight domains instead of weighting preparation toward Applications and Integration (33.1%) and Model Selection and Optimization (16.8%), which together make up just under half the exam. A candidate who studies Claude Code (3.1%) and Eval/Testing/Debugging (2.6%) as hard as the two heaviest domains is misallocating effort relative to how much each domain actually affects the scaled score.

It's reasonable to bring it into that conversation as one data point among others — a verifiable, third-party-assessed signal of Claude-integration competence. It's not a guarantee of any specific outcome, and it works best alongside a track record of actual delivered work, not as a substitute for one.

Not that's publicly referenced in the exam guide or Partner Academy materials at this time. Without a published pass rate, the most reliable way to gauge your own readiness is the eligibility self-assessment and domain-mapping approach covered on the [eligibility page](/claude-certified-developer/eligibility), rather than benchmarking against an unpublished statistic.

General LLM-integration skills — tool/function calling, context management, prompt design, handling streaming — transfer reasonably well conceptually, since the underlying mechanics are similar industry-wide. But CCDV-F's specific questions reference Claude-specific tooling (the Claude Agent SDK, Claude Code, Claude Skills, MCP as Anthropic implements it), so budget time to translate general experience into Claude-specific terms rather than assuming full transfer.

Anthropic's own exam guide describes the domain structure and task statements in detail, though it may not include full sample questions. At least one third-party paid exam-prep platform has published a gated free sample test matching this exam's structure. This site doesn't reproduce exam questions (which would require access Anthropic hasn't published), but covers the domain blueprint and realistic scenario types in depth on the [hub page](/claude-certified-developer).

It's notably more concentrated. Architect Foundations' heaviest domain (Agentic Architecture & Orchestration) is 27%, Associate Foundations' heaviest (Output Evaluation and Validation) is 21%, and Architect Professional's heaviest (Integration) is 19%. CCDV-F's 33.1% is the most skewed single-domain weighting among all four credentials, making domain-weighted preparation especially important for this specific exam.

Often, yes. At $125, the fee typically fits within existing professional-development budgets, and unlike many certifications, the CCDV-F domain blueprint maps directly onto work most engineering teams already want their Claude-focused engineers doing well, which tends to make the reimbursement conversation with a technical manager straightforward. See [cost and registration](/claude-certified-developer/cost-registration) for how to frame the ask.

The exam is delivered through Pearson VUE, which operates test centers and online proctoring internationally, and registration happens through the Anthropic Partner Academy regardless of location. The credential itself isn't region-specific — it's the same exam guide and blueprint globally, though the fee may be charged in local currency equivalent at checkout.