Anthropic launched the Claude Certification Program on March 12, 2026, alongside the Claude Partner Network, starting with one exam (Architect, Foundations) and expanding within months to four exams across three roles: Associate (for consultants, sellers, and delivery leads scoping Claude use cases), Developer (for engineers building with the Claude API, Claude Code, and Model Context Protocol), and Architect, split into Foundations and Professional levels (for partners designing end-to-end Claude deployments). All four exams run 120 minutes with a 720-out-of-1000 passing score, and delivery moved onto Pearson VUE (online proctored or test center) on June 30, 2026. By June 3, 2026, Anthropic reported over 10,000 consultants had already earned a Claude certification.
Claude adoption inside enterprises is accelerating, and Anthropic backs its Partner Network with a $100 million investment in training, technical support, and co-selling, which puts real commercial weight behind these credentials for consulting and systems-integration firms. As the only official Claude credential, it is also the one AI certification consulting buyers can point to when checking whether a vendor's staff actually know the Claude API, Agent SDK, and MCP rather than claiming generic LLM experience.
Anthropic launched the Claude Certification Program on March 12, 2026, alongside the Claude Partner Network, starting with a single exam (Architect, Foundations) before expanding to the current four-exam lineup across three roles. By June 3, 2026, Anthropic reported more than 10,000 consultants had earned a Claude certification. The lineup is Claude Certified Associate - Foundations (CCAO-F), Claude Certified Developer - Foundations (CCDV-F), Claude Certified Architect - Foundations (CCAR-F), and Claude Certified Architect - Professional (CCAR-P). All four run 120 minutes, mix multiple-choice and scenario-based multiple-response questions, and require a scaled score of 720 out of 1000 to pass. Exam delivery moved onto Pearson VUE on June 30, 2026, and you can test online proctored (OnVUE) or at a Pearson test center.
None of the four exams has a formal prerequisite, so you can sit any of them directly, but there is a real gate before that: registration needs a partner email address on a recognized company domain tied to an organization in the Claude Partner Network, and Anthropic blocks personal addresses like Gmail at checkout (its own certification FAQ notes that linking a new company domain to a partner record can take 7 to 10 business days). Joining the Partner Network itself is free for any organization actively bringing Claude to market. Once you are in, pick by role rather than working through a ladder: Associate fits consultants and sellers who scope Claude use cases, though it does not count toward Partner Network tier eligibility, Developer fits engineers shipping Claude API, Claude Code, and MCP integrations, and Architect, taken at Foundations then Professional, fits people designing production and multi-model deployments. Certifications stay valid for 12 months, and a failed attempt costs you 14 days before a retry, then 30, then 90, capped at 4 attempts per rolling year.
CertCompanion carries all four Claude practice banks: 197 questions for Associate - Foundations (CCAO-F), 302 for Developer - Foundations (CCDV-F), 314 for Architect - Foundations (CCAR-F), and 306 for Architect - Professional (CCAR-P). Every bank starts with 30 free questions, so you can gauge the exam's Claude API, Agent SDK, and MCP-heavy question style before paying for anything.
The company-email gate is the detail most guides bury: you cannot buy a Claude exam voucher with a personal Gmail account no matter how much hands-on Claude API experience you have, which makes this the only major AI vendor certification closed to individual practitioners by design. That is also likely why detailed candidate reviews are still thin on open forums as of September 2026, most of what circulates so far comes from partner-side consultants rather than the public threads other certifications get picked apart on.
Anthropic · CCAR-F
Validates foundational knowledge and skills for designing and operating production-grade agentic AI systems using Claude, including the Claude API, Agent SDK, Claude Code, and Model Context Protocol (MCP). Targeted at software engineers, solutions architects, and AI developers building real-world Claude-powered applications.
Anthropic · CCAR-P
Validates advanced expertise in designing and architecting production-grade AI solutions using Claude, targeting solution architects, AI/ML engineers, and technical leads with 3+ years of systems-architecture experience. Covers solution design, integration patterns (MCP, API, agent-to-agent), evaluation, governance, safety, and operational enablement.
Anthropic · CCAO-F
Validates foundational skills for non-technical professionals who use Claude as a productivity and workflow tool, covering responsible AI use, prompt engineering, output evaluation, and workflow integration without requiring coding or API experience.
Anthropic · CCDV-F
Validates foundational knowledge and skills for developers building applications with Anthropic's Claude AI models. Covers API integration, prompt engineering, model selection, agents, workflows, and security best practices.