Practical, up-to-date documentation for every major AI model and tool — Claude, ChatGPT/GPT, Gemini, plus core concepts like RAG, MCP, agents and prompting. Each page is concise, example-driven, and shows when it was last verified.
Read the reference, then run the examples against a live model in the AI4AI Playground. Browse by category below.
Anthropic's model family — flagship Fable 5, balanced Sonnet, fast Haiku,Strongest at coding, agents, and long documents (200K context),Use prompt caching + th…
GPT-5.x for general work; o-series for cheap chain-of-thought reasoning,Responses API is the current standard; structured outputs enforce JSON Schema,Strong mu…
Gemini 3 Pro: 2M context — the long-context champion,Flash tier is the price/speed leader for high-volume work,Native video + audio understanding; strong Works…
AI-native code editor (VS Code fork) — Tab autocomplete, inline edits, Agent mode,Agent mode executes multi-file changes and terminal commands,Rules files (.cu…
AI grounded in your work data (Microsoft Graph), across Word, Excel, PowerPoint, Outlook & Teams,Prompt with Goal + Context + Source + Expectations; name sourc…
Three layers: inline completions, Copilot Chat, and agentic modes,Multi-model — pick GPT, Claude, or Gemini variants in the chat,Teach it your codebase with .g…
Open standard connecting AI apps to tools & data — "USB for AI",Build one MCP server, use it from Claude Code, desktop, IDEs, your apps,Primitives: tools, reso…
Fetch relevant documents first, then let the model answer from them,Fixes: stale knowledge, hallucination on private data, citation needs,Modern stack: hybrid …
Concise, practical reference pages for the major AI models and tools — what they are, when to use them, and copy-pasteable examples — kept current as the tools change.
Yes, all documentation is free to read. Sign in to run the examples against live models in the playground.
Each doc shows when it was last verified, and pages are revised as providers ship new models and features.