Claude basics

Claude is Anthropic’s family of language models and the products built around them. Think of Claude as the assistant / model brand, not as a cloud. Clouds like AWS Bedrock are separate platforms that may host Claude (and many other models).

Anthropic product surfaces

Surface	What beginners notice
Chat	A conversation UI (claude.ai and similar) for drafting, explaining, Q&A.
API	Apps call Claude programmatically through Anthropic’s developer API.
Claude Code	A coding agent / CLI that works in your project folder — covered next in Claude Code basics.

You can learn Foundations habits with any chat assistant. You do not need a paid plan or API key to finish reading this path.

What “a Claude model” means

Anthropic ships different model versions over time. Treat “Claude” as a line of models: same general idea (a large language model for dialogue and text), with version differences in speed, cost, and capability. Check the product’s own docs for which version you’re on when it matters — we don’t pin version numbers here so they don’t go stale.

High-level strengths (tendencies, not scores)

People often use Claude for careful drafting, step-by-step explanations, and working with pasted documents. Those are tendencies. Any LLM can invent facts — keep verify habits.

Claude ≠ Bedrock

	Claude	Bedrock
Who makes it	Anthropic	Amazon Web Services
What it is	Models + Anthropic products	AWS service to call many foundation models
Your mental model	“The assistant / model”	“A cloud door to models”

Same model family can be reached via Anthropic’s own products or via a cloud provider. Different door, different billing and permissions — see how clouds serve frontier models.

Neutral stance

vaathi.dev is not affiliated with Anthropic. We name Claude because beginners hear it often. The same prompting and risk habits apply across assistants.

See also: Glossary — Claude.

Takeaway

Claude = Anthropic’s models and product surfaces (chat, API, Claude Code). Cloud hosting is a separate story.