What is a Large Language Model (LLM)?

A Large Language Model is a large-scale Transformer model trained on massive text corpora. Representative models include GPT, Claude, Gemini, Llama, DeepSeek, and Qwen.

An AI Agent is a system that can perceive context, plan steps, call tools, manage memory, and take actions to complete multi-step tasks.

What is RAG (Retrieval Augmented Generation)?

RAG retrieves relevant documents from a knowledge base and provides them as context to an LLM, improving grounded answers and private-data access.

What is Prompt Engineering?

Prompt engineering is the practice of designing clear, specific, and structured instructions so AI models produce more useful outputs.

What is the Transformer architecture?

The Transformer is a neural network architecture introduced by Google in 2017. Its self-attention mechanism underpins most modern foundation models.

What is MCP (Model Context Protocol)?

MCP is an open protocol for connecting AI models to external tools and data sources through a standard interface.

AI123Box · Global AI Tools Directory and Knowledge Map

AI123Box · Global AI Tools Directory and Knowledge MapAI123Box curates 3000+ AI tools across chatbots, image generation, video generation, AI agents, coding, search, research, productivity, marketing, open-source projects, and community resources. The site supports multilingual URLs, hreflang, canonical pages, RSS, OpenAPI, and agent-readable data surfaces.Popular categories

Chatbots — ChatGPT, Claude, Gemini, DeepSeek

Image generation — Midjourney, Stable Diffusion, DALL·E, FLUX

Video generation — Sora, Runway, Kling, Pika

AI agents — AutoGPT, LangChain, Claude Code, Manus

Coding — Cursor, GitHub Copilot, v0.dev

AI search — Perplexity and other answer engines

Audio and music — Suno, ElevenLabs, Udio

Productivity — Notion AI, Gamma, NotebookLM

AI knowledge map150+ AI concepts across fundamentals, models, agents, technical principles, use cases, frontier trends, and prompt engineering.

Large Language Model (LLM) — large-scale language models trained on massive text corpora

Transformer — the core architecture behind modern foundation models

RAG — retrieval augmented generation for grounded answers

AI Agent — systems that plan, use tools, and complete multi-step tasks

Prompt Engineering — designing prompts for better model output

RLHF — reinforcement learning from human feedback

MCP — Model Context Protocol for connecting AI models with tools

Scaling Law — empirical relationships between model scale, data, compute, and quality

Deep dives

Is your product AI Native? — a decision tree for evaluating real AI depth.

Is your company AI Native? — assess models, data loops, organization, revenue, and culture.

For AI agents / bots

/llms.txt — site map formatted for LLM ingestion.

/llms-full.txt — every curated tool in a single plain-text dump.

/openapi.json — OpenAPI 3.1 description of the public search API.

/.well-known/ai-plugin.json — AI plugin manifest.

/api/tools — CORS-enabled JSON search endpoint (params: search, category, sort, limit, offset, id).

/data/tools.json — full snapshot.

/feed.xml — RSS of the latest curated tools.

/mcp — MCP (Model Context Protocol) server, exposes search_tools / get_tool / list_categories.

/tool/<id> · /category/<name> — SSR HTML pages per tool / category with SoftwareApplication & CollectionPage schema.