The credentials

Every Claude certification Anthropic offers. We hold them.

The Claude Certification Program is role based: Practitioner, Developer, and Architect. Four proctored exams, each mapped to real work rather than to prompting trivia.

CCA-F

Claude Certified Associate, Foundations

Foundations

Validates fluent, effective everyday use of Claude: prompting technique, Projects and artifacts, safe handling of company data, and advising colleagues on where Claude fits.

Claude fundamentalsPrompting techniqueResponsible useWorkflow design
Practitioner · 120 minutes · $99Held on our bench
CCD-F

Claude Certified Developer, Foundations

Foundations

Validates building on the Claude Platform: the Messages API, structured output, tool use, Model Context Protocol servers, prompt caching, batch processing, and agentic workflows with Claude Code.

Claude APITool use and MCPAgent implementationCost and caching
Developer · 120 minutes · $125Held on our bench
CCAR-F

Claude Certified Architect, Foundations

Foundations

Validates designing Claude solutions end to end: use case selection, context and retrieval architecture, evaluation strategy, deployment surface, and governance for a single workload.

Solution designRetrieval architectureEvaluation strategyGovernance
Architect · 120 minutes · $125Held on our bench
Exam prep & build exercises
CCAR-P

Claude Certified Architect, Professional

Professional

The advanced tier. Validates leading enterprise-scale Claude deployments: multi-agent systems, platform choice across the Anthropic API, Amazon Bedrock and Google Vertex, security review, and organisation-wide rollout.

Enterprise rolloutMulti-agent systemsSecurity and compliancePlatform strategy
Architect · 120 minutes · $175Held on our bench
All four exams are proctored, run 120 minutes, and are delivered through Pearson VUE. Prices shown are Anthropic's list prices and can change.

Also held

Anthropic Academy course certificates

Beyond the proctored exams, our engineers carry the full Anthropic Academy track. It is how we stay current as Claude ships new capabilities.

Claude 101

Core Claude usage and features

Claude Code 101

Agentic coding in daily development

Claude Platform 101

Building on the API from the ground up

Model Context Protocol

Designing and shipping MCP servers

AI Fluency

Framework for effective, responsible AI work

Claude Code in action

Advanced repo and workflow patterns

Introduction to Cowork

Collaborative agent workflows

Prompt engineering

Structured prompting and long context

The exam domains

What a Claude Certified Engineer is actually tested on.

Six disciplines, tested rather than claimed. Every engineer we staff holds the full set, and every line of code they write is reviewed by a second certified engineer.

01

Prompt and context engineering

Structured prompting, XML scaffolding, system-prompt design, long-context strategy, and prompt caching that cuts spend without cutting quality.

Tested in: Associate and Developer

02

Tool use and MCP

Designing tool schemas Claude actually calls correctly, building Model Context Protocol servers, and wiring them into internal systems safely.

Tested in: Developer

03

Agent architecture

Multi-step agents with planning, sub-agents, memory, retries, budget ceilings, and human-in-the-loop checkpoints at the right seams.

Tested in: Developer and Architect

04

Evaluation and quality

Golden datasets, LLM-as-judge rubrics, regression gates in CI, and dashboards that make model quality a number your team can defend.

Tested in: Architect, Foundations

05

Safety and governance

Refusal handling, injection resistance, PII redaction, audit trails, and deployment patterns that clear security and legal review.

Tested in: Architect, Professional

06

Claude Code and developer velocity

Agentic coding workflows, repo conventions, CLAUDE.md design, and enablement so your own engineers keep shipping after we leave.

Tested in: Developer