Anthropic Claude¶

This work is licensed under a Creative Commons Attribution 4.0 International License.

Ways to Access Claude¶

There are multiple ways to access Claude:

1. Claude Chat Interface (claude.ai):

• Go to: https://claude.ai/
• Sign up: Create an account using your email address or with a Google account
• Log in: If you already have an account, log in with your credentials

2. Claude Code (terminal, VS Code/JetBrains, desktop app, and web):

• Install: curl -fsSL https://claude.ai/install.sh | bash (recommended) or npm install -g @anthropic-ai/claude-code, then run claude in your project folder; also available as a VS Code / JetBrains extension, in the Claude desktop app, and on the web at claude.ai/code. Docs: code.claude.com/docs
• Features: Agentic pair programming — reads your codebase, writes and refactors code across files, runs commands, and manages Git
• Authentication: Included with Pro, Max, Team, and Enterprise plans, or use an API key

3. Claude Desktop App:

• Download: Available for macOS and Windows at claude.ai/download
• Features: Native desktop experience with keyboard shortcuts, file handling, and system integration
• Model Context Protocol: Built-in MCP support for connecting to local tools and services

4. Anthropic API (for Developers):

• Sign Up: Go to https://console.anthropic.com/ to create an account
• API Key: Generate an API key from your console dashboard
• Documentation: https://docs.claude.com/

Model Context Protocol (MCP)¶

The Model Context Protocol is an open standard that enables Claude to interact with external tools and data sources:

What is MCP?

• Purpose: Allows Claude to connect to databases, APIs, files, and other tools on your computer
• Security: Runs locally with your explicit permission for each connection
• Open Standard: Developed by Anthropic and available as open-source

Installing MCP:

1. For Claude Desktop:

2. MCP support is built into Claude Desktop

3. Configure servers in Settings → Developer → Model Context Protocol

4. Add server configurations in JSON format

5. Example MCP Configuration:

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/allowed/directory"]
    },
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_PERSONAL_ACCESS_TOKEN": "your-token-here"
      }
    }
  }
}

1. Popular MCP Servers:
2. Filesystem: Access local files and directories
3. GitHub: Interact with GitHub repositories
4. PostgreSQL: Query databases
5. Slack: Read Slack messages
6. Google Drive: Access Google Drive files

Learn More: modelcontextprotocol.io

Using Claude¶

Web Chat Interface (claude.ai):

• Prompting: Type your requests or questions into the chat box. Be clear and specific in your prompts
• Conversation History: Claude remembers the context of your conversation within the current chat
• Projects: Organize chats into projects with custom instructions and shared knowledge
• Artifacts: Claude can create and edit code, documents, and diagrams in a dedicated panel
• File Uploads: Upload images, PDFs, and text files (up to 5 files, 10MB each)

Claude Code (terminal, VS Code/JetBrains, desktop app, and web):

• Installation:
  1. Run curl -fsSL https://claude.ai/install.sh | bash (recommended) or npm install -g @anthropic-ai/claude-code
  2. Open a terminal in your project folder and run claude
  3. Optionally install the "Claude Code" extension for VS Code or JetBrains, or use it from the Claude desktop app or claude.ai/code
  4. Sign in with your Claude account (Pro, Max, Team, or Enterprise) or an API key
• Features:
  • Agentic coding: reads, writes, and refactors code across multiple files
  • Runs tests and shell commands and manages Git through conversation
  • Chat panel and inline edits in VS Code and JetBrains
  • Multi-file context awareness
  • Full walkthrough: Claude Code tutorial and code.claude.com/docs

Claude Desktop App:

• Installation:
  • macOS: Download from claude.ai/download and drag to Applications
  • Windows: Download installer and follow setup wizard
• Features:
  • Native OS integration
  • Global keyboard shortcuts
  • MCP server connections
  • Local file access (with permission)
  • Offline viewing of past conversations

Anthropic API:

• Quick Start (Python):

  from anthropic import Anthropic
  
  client = Anthropic(api_key="your-api-key")
  
  response = client.messages.create(
      model="claude-sonnet-5-5",  # current IDs: https://platform.claude.com/docs/en/about-claude/models/overview
      max_tokens=1000,
      messages=[
          {"role": "user", "content": "Hello, Claude!"}
      ]
  )
  print(response.content[0].text)
  

• SDKs Available: Python, TypeScript/JavaScript, Go, and community SDKs
• Use Cases: Chatbots, content generation, code assistance, data analysis

Tips for Using Claude¶

• Be Specific: Provide clear instructions and context in your prompts.
• Iterate: Refine your prompts based on Claude's responses to improve the results.
• Use System Prompts: For complex or multi-step tasks, consider using system prompts to provide overall instructions to guide Claude's behavior.
• Experiment: Try different prompting techniques and model settings to find what works best for your use case.

About Claude¶

Claude is a family of large language models (LLMs) developed by Anthropic, a company focused on AI safety and research. Claude is known for:

• Helpful and Honest Responses: Designed with Constitutional AI for safer, more aligned outputs
• Advanced Reasoning: Excels at complex analysis, math, and multi-step problem-solving
• Strong Coding Abilities: Excellent for software development, debugging, and code review
• Large Context Window: 1M tokens (about 750,000 words) on current models
• Vision Capabilities: Can analyze images, charts, diagrams, and screenshots

Claude Model Family¶

Anthropic publishes Claude in tiers. As of September 2026 the current lineup is Fable 5.1, Opus 5.5, Sonnet 5.5, and Haiku 4.5. For an authoritative, up-to-date list of model IDs, see the Claude models overview.

• Fable (Claude Fable 5.1):

  • Frontier tier — Anthropic's most capable model, for the hardest reasoning and long-running agentic tasks
  • Model ID: claude-fable-5-1

• Opus (Claude Opus 5.5):

  • Flagship tier — complex reasoning and advanced coding at a lower price than Fable
  • Model ID: claude-opus-5-5

• Sonnet (Claude Sonnet 5.5):

  • Balanced tier — best for everyday coding, analysis, and creative tasks, with an excellent performance-to-cost ratio
  • Model ID: claude-sonnet-5-5

• Haiku (Claude Haiku 4.5):

  • Fast, cost-effective tier — great for simple tasks and high-volume applications
  • Model ID: claude-haiku-4-5

Further Resources¶

• Anthropic Website: https://www.anthropic.com/
• Claude Documentation: https://docs.claude.com/
• API Reference: https://docs.claude.com/en/api/
• Prompt Engineering Guide: https://docs.claude.com/en/docs/build-with-claude/prompt-engineering
• Claude Code Documentation: https://code.claude.com/docs/en/overview
• Pricing: https://claude.com/pricing
• Model Context Protocol: https://modelcontextprotocol.io
• Anthropic Cookbook: https://github.com/anthropics/anthropic-cookbook
• Community Discord: https://discord.gg/anthropic