Anthropic's Claude isn't one product; it's five separate layers that most enterprise buyers confuse with each other. Ask three people at the same company what "Claude" means, and you'll likely get three different answers: a chatbot, a coding tool, or an API line item. Understanding the distinction between the model itself, how you access it, the applications built on top, the protocol connecting Claude to your tools, and where it all runs is crucial before any enterprise rollout.

The most common mistake in deploying Claude happens when teams buy the wrong layer for their actual problem. A team that purchases Claude Pro seats gets chat, Cowork, and Claude Code for individuals, but a team needing Claude wired into an internal ticketing system needs the API and the Model Context Protocol (MCP) layer instead, regardless of how many chat seats it already owns.

What Are the Five Layers of Anthropic's Stack?

The Anthropic AI stack consists of five distinct layers that sit between a raw Claude model and the actual work someone performs. Each layer solves a different problem, and conflating them is where most procurement conversations go wrong.

• The Model Layer: Anthropic's current lineup includes Claude Haiku 4.5 for fast, low-cost tasks; Claude Sonnet 5 for most coding, agent, and enterprise workflows; Claude Opus 5 for complex agentic coding and high-stakes reasoning; and Claude Fable 5 as the most generally capable model. Above Fable 5 sits Claude Mythos 5, the same underlying model without Fable's extra safety filtering, available only to approved partners doing defensive cybersecurity work.
• The Access Method Layer: Two separate paths lead to the same models. Claude.ai subscription plans (Free, Pro, Max, Team, and Enterprise) are billed by seat and cover chat, Claude Code, and Claude Cowork. The Claude Platform, or API, bills by token instead and is for engineering teams building custom software, with options for the raw Messages API or Claude Managed Agents, a hosted agent harness that Anthropic runs for you.
• The Application Layer: Claude Code targets software engineering, writing, debugging, testing, and shipping code. Claude Cowork applies the same read-plan-execute loop to non-coding knowledge work such as research, file organization, and document production. Additional surfaces include Claude in Chrome for browser automation and Claude in Excel and PowerPoint for spreadsheet and slide work.
• The Protocol Layer: MCP, the Model Context Protocol, is the open standard that Anthropic introduced and open-sourced in November 2024 for connecting any AI application to any external tool or data source through a single shared interface. This turns what used to be an M times N integration problem into an M plus N problem, reducing custom integrations significantly.
• The Deployment Layer: The same Claude models are available via five deployment routes: Anthropic's own Claude API, Amazon Bedrock, Claude Platform on AWS billed through AWS Marketplace, Google Cloud, and Microsoft Foundry.

How Should Teams Choose the Right Claude Layer for Their Needs?

Anthropic's own guidance for most enterprise coding and workflow automation is straightforward: start on Sonnet 5, move to Opus 5 when a task's stakes or complexity justify the extra cost, and reserve Haiku 4.5 for high-volume, low-complexity work like support-ticket triage. Pricing scales with capability, from $1 input and $5 output per million tokens on Haiku 4.5 up to $10 and $50 on Fable 5.

On the subscription side, Pro runs roughly $20 per month (about $17 billed annually) and is the entry point that includes Claude Code and Cowork. Max adds 5x or 20x the usage headroom at $100 or $200 per month for heavy individual users. Team starts around $20 to $25 per standard seat with a five-seat minimum, with premium seats near $100 to $125 adding Claude Code. Enterprise layers on single sign-on (SSO), automated user provisioning (SCIM), and compliance controls at custom pricing.

On the API side, the Messages API is stateless, meaning it holds no memory of earlier turns unless you resend them, and you own the agent loop yourself. Claude Managed Agents, in public beta since April 2026, handles sandboxing, isolated execution, state persistence, and error recovery for you at $0.08 per session-runtime hour plus standard token costs, with early adopters including Notion and Asana.

Why Does the MCP Protocol Matter for Enterprise Deployments?

MCP is what lets Claude Code, Claude Cowork, Claude Desktop chat, and custom Agent SDK builds all reach the same connector catalog instead of needing four separate integrations. Every Claude.ai plan supports connecting MCP servers to Claude Desktop, and OpenAI, Microsoft, and Google DeepMind have all adopted the same protocol since Anthropic released it.

This matters significantly for procurement, since an MCP server your team builds for one AI product is very likely to work with others later. For teams that need something the packaged apps don't offer, the Agent SDK sits one layer below Claude Code and Cowork, giving developers the same harness, tool ecosystem, and MCP connectivity to build a fully custom agent from scratch.

All models on Claude 4.6 and later, including Fable 5, ship with a 1-million-token context window as standard rather than an opt-in add-on, allowing them to process roughly 100,000 words at once. This capability matters for teams working with large documents, codebases, or research materials that require sustained context.

Understanding which layer solves which problem is the foundation of any successful Claude deployment. Rather than asking "Should we buy Claude?" teams should ask "Which layer of the Anthropic stack solves our specific problem?" The answer determines not just cost, but whether the deployment succeeds at all.