Home / Glossary
Glossary
Plain-language definitions of common AI automation terms: LLM, token, API key, agent, MCP, LoRA and more.
- Large language model (LLM)
- An AI model trained on large amounts of text that can understand and generate natural language. ChatGPT, Claude and Gemini are all built on models like this. → LLM API
- Prompt
- The instructions and background you give an AI. The more specific it is, the more consistent the result tends to be. → Start Here
- Hallucination
- When a model states something incorrect or made up with full confidence. Always verify important facts yourself. → Start Here
- API
- An interface that lets programs talk to each other. Calling an LLM API means sending a prompt to the model from code and getting its reply back. → LLM API
- API key
- A secret string that identifies you and authorizes your use of an API. Treat it like a password and never paste it into public code, screenshots or videos. → LLM API
- Token
- The smallest unit of text a model processes: a word, part of a word, or a symbol. APIs bill by token count, and context length is measured in tokens too. → LLM API
- Context window
- The maximum amount of content, measured in tokens, a model can see at once, covering both your input and its reply. Anything beyond it has to be trimmed or handled separately. → LLM API
- Tool calling
- Instead of answering directly, the model asks your code to run a function (look something up, write a file, send a request) and then continues with the result. → LLM API
- RAG (retrieval-augmented generation)
- Find the relevant passages in your own documents first, then hand them to the model together with the question, so the answer is grounded in your data rather than only the model's memory. → LLM API
- AI agent
- An AI system that breaks a goal into steps, picks tools, acts step by step and adjusts based on the results, rather than just answering one question at a time. → AI Agents
- MCP (Model Context Protocol)
- An open protocol that gives AI applications a standard way to connect to external tools and data sources. → AI Agents
- Workflow
- A task broken into fixed steps in a fixed order. In ComfyUI it is a graph of connected nodes; in automation it is a sequence that runs by itself. → Automation
- Webhook
- An HTTP notification a system sends on its own when an event happens. Often used to trigger an automation. → Automation
- Node
- A building block in ComfyUI with inputs and outputs. Wires connect nodes into a workflow, and a busy graph looks like a bowl of noodles, hence the community nickname. → ComfyUI
- Custom node
- A community-made ComfyUI node you install separately to add features ComfyUI does not have out of the box. → ComfyUI
- ComfyUI Manager
- A tool for installing, updating and managing custom nodes and models. It is now built into ComfyUI; on the portable package you switch it on with the --enable-manager flag. → ComfyUI
- Checkpoint
- A file holding a model's full weights. It sets the basic style and capability of the images you generate. → ComfyUI