AI GLOSSARY

28 AI terms explained the way you would explain them to a smart friend: what it is, why it matters, a concrete example, and the most common misunderstanding.

Zero retention: everything you type stays in this browser. Verify it yourself in DevTools.
Guide

What Do These AI Terms Mean?

Agent instructions file
A Markdown file in your repository that tells coding agents your commands, conventions and rules.
Agent loop
The repeated cycle of decide, act with a tool, observe the result, and decide again that makes an AI an agent.
Agentic workflow
A setup where an AI takes multiple steps on its own, using tools and checking results, rather than giving one reply.
Context engineering
Deciding what goes into a model context window, including instructions, files and tool output, so the model can do the task well.
Context window
The maximum amount of text, measured in tokens, that a model can take into account in a single request.
Distillation
Training a small model to imitate a larger model, so you get much of the quality at a fraction of the cost.
Embedding
A list of numbers that represents the meaning of text, so similar texts can be found by comparing numbers.
Eval
A repeatable test set with a scoring method, used to measure whether an AI system does its job and catch regressions.
Fine-tuning
Extra training on your own examples that adjusts a model weights so it handles a specific task or style better.
Frontier model
One of the most capable AI models at a given moment, typically closed and served by a large lab through an API.
Guardrail
A rule or check, in the instructions or in code, that keeps an AI system from doing what it should not.
Hallucination
When a model confidently produces something false or made up, such as an API that does not exist.
Inference
Running an already trained model to produce output. Every reply you get from an AI is one inference call.
Large language model
An AI model trained on vast amounts of text to predict the next token, which lets it write, answer questions and code.
MCP (Model Context Protocol)
An open standard for connecting AI applications to external tools and data sources through a common interface.
Mixture of experts
A model design that activates only a few specialist sub-networks per token, making huge models cheaper to run.
Open-weight model
A model whose trained weights you can download and run on your own hardware, under the terms of its license.
Prompt caching
Reusing the already processed start of a prompt across requests so repeated context is cheaper and faster.
Prompt injection
An attack where instructions hidden in content an AI reads trick it into acting against its user.
Quantization
Storing model weights with fewer bits so a model needs less memory and runs faster, with some loss of quality.
RAG (retrieval-augmented generation)
Looking up relevant documents first and giving them to the model, so its answer is grounded in real sources.
Reasoning model
A model trained to think through a problem in steps before answering, using more time and tokens for harder tasks.
Recursive self-improvement
The idea of AI systems helping to build better AI, which then builds better AI again. Partly real, mostly still speculative.
Subagent
A helper agent that a main agent starts for one focused task, with its own separate context, returning a summary.
System prompt
The standing instructions an application gives a model before the conversation, setting its role, rules and format.
Temperature
A setting that controls how random or predictable a model output is when it picks each next token.
Token
A small chunk of text, often part of a word, that models process; limits and prices are counted in tokens.
Tool use
A model asking its application to run a defined function, such as search or run tests, and using the result.

Why Another AI Glossary?

Most AI glossaries are either academic or marketing. This one is written for people who use AI tools at work, especially coding agents, and want to know what a term means in practice. Every entry has an honest tone: where something is speculative, it says so, and where a common claim is wrong, it says that too.

  • One plain paragraph first, details after.
  • A concrete example on every page.
  • No benchmark numbers that go stale in a month.

Where Should You Start?

  1. Read token and context window: almost everything else builds on them.
  2. Then agent loop and tool use, which explain what coding agents actually do.
  3. Then context engineering and agent instructions file, which explain why files like AGENTS.md matter.

For a plain comparison of the models themselves, see frontier vs open-weight models.

Frequently Asked Questions

Who writes the glossary?

The 2prompt team, from vendor documentation and common usage. Definitions are kept short and checked for accuracy, and we avoid claims about specific benchmark numbers.

Is the glossary free to quote?

Yes. Quote or link any definition. A link back is appreciated but not required.

Why are some definitions cautious?

Because some terms, such as recursive self-improvement, mix real current practice with speculation. The definition separates the two instead of picking a side.

Can I suggest a term?

Yes. Use the feedback buttons at the bottom of any page and name the term you would like explained.

Was this helpful?