Confused by AI terms like prompts, tokens, LLMs, RAG, hallucinations, embeddings, agents, and multimodal AI? This beginner-friendly AI glossary explains 100+ artificial intelligence terms in simple language.
Artificial intelligence has introduced hundreds of new terms, concepts, tools, and technologies.
Whether you’re learning how to use ChatGPT, Gemini, Claude, AI image generators, AI video tools, or other artificial intelligence systems, understanding the terminology can make it much easier to know what you’re doing.
This AI Glossary explains more than 100 important artificial intelligence terms in simple, beginner-friendly language.
You’ll find definitions for common AI terms, including AI prompts, prompt engineering, large language models, machine learning, generative AI, tokens, context windows, embeddings, RAG, AI agents, hallucinations, fine-tuning, multimodal AI, APIs, and more.
If you’re new to prompting, start with our Prompting for Beginners guide.
You can also explore our Prompt Anatomy guide to understand the different parts of an effective AI prompt.
What Is Artificial Intelligence?
Artificial intelligence, commonly called AI, refers to computer systems designed to perform tasks that normally require human intelligence.
These tasks can include:
- Understanding language
- Recognizing images
- Generating text
- Creating images
- Producing videos
- Analyzing information
- Recognizing patterns
- Making predictions
- Solving problems
- Assisting with decisions
Modern AI systems can perform many of these tasks using machine learning and large datasets.
AI and Machine Learning Terms
Artificial Intelligence (AI)
Artificial intelligence is the broader field of creating computer systems capable of performing tasks associated with human intelligence.
Machine Learning (ML)
Machine learning is a branch of AI where systems learn patterns from data instead of being explicitly programmed for every individual task.
Deep Learning
Deep learning is a type of machine learning that uses neural networks with multiple layers to process complex patterns in data.
Neural Network
A neural network is a machine-learning model inspired loosely by the way biological brains process information. It consists of connected computational units that learn patterns from data.
Training Data
Training data is the information used to train a machine-learning model.
The quality, quantity, and characteristics of training data can significantly influence how a model performs.
Dataset
A dataset is a collection of information used for training, evaluating, or testing an AI system.
Model
An AI model is a trained computational system that can process inputs and produce outputs.
Algorithm
An algorithm is a set of instructions or procedures used to solve a problem or perform a computation.
Inference
Inference is the process of using a trained AI model to generate an output from new input.
For example, when you send a prompt to an AI chatbot and receive an answer, the model is performing inference.
Neural Network Layer
A layer is a processing stage within a neural network. Different layers transform information as it moves through the model.
Parameter
A parameter is a value learned by an AI model during training. Modern AI models can contain extremely large numbers of parameters.
Weight
A weight is a learned value within a neural network that influences how strongly different pieces of information affect the model’s output.
Generative AI Terms
Generative AI
Generative AI refers to AI systems capable of creating new content based on patterns learned from data.
Generative AI can produce:
- Text
- Images
- Audio
- Video
- Code
- Music
- Presentations
AI Content Generation
AI content generation is the use of AI systems to create content from instructions or other inputs.
Text-to-Text AI
Text-to-text AI systems accept text instructions and generate text responses.
Chatbots and writing assistants are common examples.
Text-to-Image
Text-to-image systems generate images based on written descriptions.
For example:
A futuristic city at sunset with flying vehicles, cinematic lighting and a wide-angle view.
Text-to-Video
Text-to-video systems generate video content based on written prompts.
Learn more in our AI Video Prompting guide.
Image-to-Image
Image-to-image generation uses an existing image as an input and generates a modified or transformed image.
Image-to-Video
Image-to-video systems use an image as a starting point for generating video.
Multimodal AI
Multimodal AI can process or generate more than one type of information, such as text, images, audio, video, or other data.
Synthetic Data
Synthetic data is artificially generated data designed to resemble real-world data.
It can be used for training, testing, or evaluating AI systems.
Large Language Model Terms
Large Language Model (LLM)
A large language model is an AI model trained on large amounts of text and other data to understand and generate language.
LLMs power many modern AI assistants and applications.
Foundation Model
A foundation model is a broadly trained AI model that can serve as the base for many different applications and tasks.
Language Model
A language model predicts or generates language based on patterns learned from data.
Transformer
A transformer is a neural-network architecture that has become foundational to many modern language and generative AI models.
Transformers are particularly effective at processing relationships between elements in sequences such as words or tokens.
Attention
Attention is a mechanism that allows an AI model to determine which parts of an input are more relevant when processing information.
Self-Attention
Self-attention allows a model to evaluate relationships between different parts of the same input sequence.
Pretraining
Pretraining is the initial process of training a model on a large dataset before it is adapted for particular tasks.
Post-Training
Post-training refers to additional training or optimization performed after the initial pretraining stage.
Fine-Tuning
Fine-tuning involves further training an existing AI model on a more specific dataset or task to adapt its behavior.
Instruction Tuning
Instruction tuning trains a model to better follow natural-language instructions.
Alignment
AI alignment refers broadly to efforts to make AI systems behave in ways that are consistent with intended goals, values, or instructions.
AI Prompting Terms
Prompt
A prompt is the instruction, question, or information provided to an AI system to guide its response.
Prompting
Prompting is the practice of communicating with an AI model through instructions or inputs to produce a desired result.
Prompt Engineering
Prompt engineering is the systematic design and refinement of prompts to improve AI outputs.
Read our complete Prompt Engineering guide.
Prompt Template
A prompt template is a reusable prompt structure containing placeholders that can be customized for different tasks.
Prompt Library
A prompt library is a collection of reusable prompts organized by task, category, tool, or purpose.
Explore the Prompt Library for ready-to-use AI prompts.
Prompt Chaining
Prompt chaining involves connecting multiple prompts or AI tasks together so that the output from one stage becomes an input for another.
Zero-Shot Prompting
Zero-shot prompting asks an AI model to perform a task without providing examples.
One-Shot Prompting
One-shot prompting provides one example to demonstrate the desired task or output.
Few-Shot Prompting
Few-shot prompting provides several examples before asking the AI to perform a similar task.
Role Prompting
Role prompting gives an AI a specific role or perspective, such as teacher, editor, researcher, or programmer.
System Prompt
A system prompt is a set of instructions that establishes how an AI model should behave or operate within a particular application.
User Prompt
A user prompt is the instruction or request provided by the person interacting with an AI system.
Context
Context is the relevant information supplied to an AI model to help it understand a task.
Prompt Injection
Prompt injection is a technique or attack where someone attempts to manipulate an AI system into ignoring or overriding its intended instructions.
Prompt Optimization
Prompt optimization is the process of testing and refining prompts to improve their results.
Learn more about practical AI prompting techniques.
AI Token and Context Terms
Token
A token is a unit of text processed by an AI language model.
A token may represent a whole word, part of a word, punctuation, or another piece of text.
Tokenization
Tokenization is the process of breaking text into tokens that an AI model can process.
Context Window
A context window is the amount of information an AI model can consider within a particular interaction.
A larger context window allows a model to work with more information at once, although larger context does not automatically guarantee better results.
Context Length
Context length refers to the amount of information a model can process as context.
Context Management
Context management involves controlling the information provided to an AI model so that important details remain available and irrelevant information is minimized.
AI Knowledge and Retrieval Terms
Knowledge Base
A knowledge base is an organized collection of information that an AI application can use to answer questions or perform tasks.
Retrieval-Augmented Generation (RAG)
RAG combines information retrieval with AI generation.
Instead of relying only on what a model learned during training, a RAG system can retrieve relevant information from an external knowledge source and use it to generate a response.
Retrieval
Retrieval is the process of finding relevant information from a database, document collection, search index, or other information source.
Embedding
An embedding is a numerical representation of information, such as text, that captures relationships and meaning in a form that AI systems can process.
Vector
A vector is a numerical representation of information.
AI applications often use vectors to represent text, images, or other data.
Vector Database
A vector database stores and searches numerical representations called vectors.
Vector databases are commonly used in retrieval systems and AI applications.
Semantic Search
Semantic search attempts to find information based on meaning rather than relying only on exact keyword matches.
Retrieval System
A retrieval system finds relevant information that can be supplied to an AI model.
Grounding
Grounding involves connecting an AI response to specific external information or data sources.
AI Output Terms
AI Hallucination
An AI hallucination occurs when an AI system generates information that appears plausible but is incorrect, unsupported, or fabricated.
Hallucinations are an important reason to verify important AI-generated information.
Factuality
Factuality refers to how accurately an AI-generated response represents factual information.
Accuracy
Accuracy describes how closely an AI output matches the correct or intended answer.
Relevance
Relevance describes how closely an AI response addresses the user’s actual request.
Consistency
Consistency refers to how reliably an AI system produces similar or compatible results when given similar inputs.
Deterministic Output
A deterministic system produces the same result when given the same conditions and inputs.
Many AI generation systems are not fully deterministic.
Creativity
In generative AI, creativity generally refers to the ability to produce varied or novel outputs.
AI Bias
AI bias occurs when an AI system produces systematically skewed or unfair results.
Bias can arise from training data, model design, evaluation methods, or how a system is used.
AI Agent Terms
AI Agent
An AI agent is a system that can use AI models to pursue goals, make decisions, use tools, and perform multiple steps toward completing a task.
Agentic AI
Agentic AI refers to AI systems designed to perform tasks with a degree of autonomy rather than simply responding to individual prompts.
Tool Use
Tool use occurs when an AI system can interact with external tools, applications, databases, APIs, or other systems.
Function Calling
Function calling allows an AI model to request that a software application execute a predefined function.
Workflow
An AI workflow is a sequence of steps used to accomplish a task using AI and potentially other tools.
Automation
AI automation uses AI systems to perform tasks that would otherwise require manual work.
Autonomous AI
Autonomous AI refers to systems designed to operate with limited human intervention within defined boundaries.
AI Chatbot Terms
AI Chatbot
An AI chatbot is software that uses AI to communicate with users through natural-language conversations.
AI Assistant
An AI assistant is an AI-powered system designed to help users perform tasks, answer questions, generate content, or access information.
Conversational AI
Conversational AI refers to technologies that allow computers to interact with people using natural language.
Chat History
Chat history is the record of previous messages in an AI conversation that may be available to the system as context.
Memory
AI memory refers to mechanisms that allow an AI application to retain selected information across interactions.
The exact capabilities of memory vary by product.
Personalization
AI personalization involves adapting an AI system’s responses or behavior based on user preferences, context, or other information.
AI Model and Performance Terms
Model Size
Model size generally refers to the scale of an AI model, often described using the number of parameters.
Latency
Latency is the amount of time between an input being submitted and an output being received.
Throughput
Throughput refers to how much data or how many requests an AI system can process within a given period.
Benchmark
A benchmark is a standardized test or evaluation used to compare AI model performance.
Evaluation
AI evaluation is the process of measuring how well an AI system performs according to defined criteria.
Model Evaluation
Model evaluation tests an AI system’s performance on specific tasks or datasets.
Leaderboard
An AI leaderboard ranks models or systems according to selected benchmarks or evaluation criteria.
Reasoning Model
A reasoning model is an AI model designed or optimized for tasks that require more complex multi-step problem solving.
AI Safety and Responsible AI Terms
AI Safety
AI safety is the field concerned with reducing harmful or unintended behavior from AI systems.
Responsible AI
Responsible AI involves developing and using AI in ways that consider safety, fairness, transparency, privacy, accountability, and other potential impacts.
AI Ethics
AI ethics examines the ethical questions and consequences associated with artificial intelligence.
Explainable AI
Explainable AI refers to approaches that help people understand how or why an AI system produced a particular result.
Transparency
AI transparency involves providing information about how an AI system works, how it was developed, or how it is being used.
Human-in-the-Loop
Human-in-the-loop systems include human involvement in an AI workflow, such as reviewing, approving, correcting, or supervising AI outputs.
Guardrail
An AI guardrail is a rule, control, or technical mechanism designed to keep an AI system within acceptable boundaries.
Content Filter
A content filter is a mechanism that identifies, blocks, modifies, or flags certain types of content.
Privacy
AI privacy concerns how personal or sensitive information is collected, processed, stored, and used by AI systems.
AI Development Terms
API
An API, or Application Programming Interface, allows software applications to communicate with each other.
AI providers commonly offer APIs that developers can use to integrate AI models into websites and applications.
SDK
An SDK, or Software Development Kit, is a collection of tools and resources that helps developers build applications for a particular platform or service.
AI Application
An AI application is software that uses artificial intelligence to perform a particular task or provide a service.
AI Model API
An AI model API allows developers to send requests to an AI model and receive generated outputs programmatically.
Open Source AI
Open-source AI generally refers to AI models, software, code, or related resources released with permissions that allow varying degrees of public access, use, modification, or distribution.
The exact license determines what users are permitted to do.
Model Weights
Model weights are learned numerical values that represent patterns acquired during model training.
Inference API
An inference API provides programmatic access to an AI model for generating outputs from inputs.
AI Image Terms
AI Image Generator
An AI image generator is a system capable of creating images from text prompts, images, or other inputs.
Image Prompt
An image prompt is a written instruction describing the image you want an AI image generator to create.
Learn more with our AI Image Prompting guide.
Negative Prompt
A negative prompt specifies elements that should be avoided when generating an image.
Support for negative prompts varies between AI image-generation systems.
Image-to-Image Generation
Image-to-image generation uses an existing image as input to create a new or modified image.
Inpainting
Inpainting involves modifying or generating a selected area of an existing image.
Outpainting
Outpainting extends an image beyond its original boundaries by generating additional content.
Image Upscaling
Image upscaling increases the resolution or apparent detail of an image using specialized software or AI models.
Aspect Ratio
Aspect ratio describes the proportional relationship between an image’s width and height.
Common ratios include:
- 1:1
- 4:3
- 16:9
- 9:16
AI Video Terms
AI Video Generator
An AI video generator is a system that creates or modifies video using AI.
Video Prompt
A video prompt describes the scene, subject, movement, camera behavior, style, environment, or other details an AI video generator should produce.
Text-to-Video
Text-to-video systems generate video based on written instructions.
Image-to-Video
Image-to-video systems use an image as a starting point for generating video.
Camera Movement
Camera movement describes how the virtual camera moves during an AI-generated scene.
Examples include:
- Pan
- Tilt
- Zoom
- Push-in
- Pull-out
- Tracking shot
- Orbit
- Crane movement
Motion Prompting
Motion prompting involves describing how subjects, objects, environments, or cameras should move within an AI-generated video.
Video-to-Video
Video-to-video systems transform or modify existing video using AI.
AI Audio and Voice Terms
AI Voice Generator
An AI voice generator creates synthetic speech from text or other inputs.
Text-to-Speech (TTS)
Text-to-speech converts written text into spoken audio.
Speech-to-Text (STT)
Speech-to-text converts spoken language into written text.
Voice Cloning
Voice cloning uses AI to generate speech that resembles a particular voice.
The use of someone’s voice may involve important consent, privacy, and legal considerations.
AI Music Generation
AI music generation uses AI systems to create music or musical elements from prompts, examples, or other inputs.
Audio Generation
Audio generation refers to AI systems that create speech, sound effects, music, or other audio content.
AI Coding Terms
AI Coding Assistant
An AI coding assistant helps developers write, explain, debug, or modify software code.
Code Generation
Code generation uses AI to create programming code based on instructions or other inputs.
Code Completion
Code completion predicts or suggests code while a developer is writing a program.
Debugging
Debugging is the process of finding and fixing problems in software.
AI systems can assist with debugging by analyzing code and suggesting possible fixes.
Refactoring
Refactoring means restructuring existing code to improve its organization, readability, maintainability, or efficiency without changing its intended behavior.
Code Review
Code review involves examining source code to identify errors, security issues, performance problems, or opportunities for improvement.
AI Search and Information Terms
AI Search
AI search uses artificial intelligence to interpret queries and provide information, summaries, answers, or recommendations.
AI Overview
An AI overview is an AI-generated summary or answer presented in a search experience.
Retrieval
Retrieval is the process of finding relevant information from an external source.
Search-Augmented AI
Search-augmented AI systems combine AI generation with information retrieved from search engines or other information sources.
Grounded Answer
A grounded answer is an AI response based on specific information provided by a trusted source, database, document, or retrieval system.
Common AI Abbreviations
AI
Artificial Intelligence
Technology that enables computer systems to perform tasks associated with human intelligence.
ML
Machine Learning
A branch of AI where systems learn patterns from data.
DL
Deep Learning
A form of machine learning based on multi-layer neural networks.
LLM
Large Language Model
An AI model trained to understand and generate language.
NLP
Natural Language Processing
The field of AI focused on enabling computers to process and understand human language.
RAG
Retrieval-Augmented Generation
A method that combines information retrieval with AI-generated responses.
API
Application Programming Interface
A method that allows software applications to communicate with each other.
AGI
Artificial General Intelligence
A hypothetical form of AI capable of performing a broad range of intellectual tasks at a level comparable to or beyond humans.
AGI does not have one universally accepted definition.
TTS
Text-to-Speech
Technology that converts written text into spoken audio.
STT
Speech-to-Text
Technology that converts spoken language into written text.
NLP
Natural Language Processing
Technology and research focused on enabling computers to work with human language.
RLHF
Reinforcement Learning from Human Feedback
A training approach that uses human feedback to help optimize an AI model’s behavior.
VLM
Vision-Language Model
An AI model capable of processing both visual information and language.
ASR
Automatic Speech Recognition
Technology that converts spoken language into text.
AI Terms Every Beginner Should Know
If you’re completely new to artificial intelligence, start with these terms:
- Artificial Intelligence
- Machine Learning
- Generative AI
- Large Language Model
- Prompt
- Prompt Engineering
- Token
- Context Window
- Hallucination
- Fine-Tuning
- RAG
- Embedding
- AI Agent
- Multimodal AI
- API
- AI Model
- Training Data
- Inference
- AI Assistant
- AI Workflow
Understanding these concepts gives you a strong foundation for learning how modern AI tools work.
AI Prompting Terms to Learn Next
If your main interest is learning how to get better results from AI, focus on these concepts:
Start With the Basics
Learn:
- Prompt
- Context
- Instruction
- Role
- Output format
- Constraints
- Examples
Our Prompt Anatomy guide explains how these components work together.
Move Into Prompt Engineering
Next, learn:
- Prompt engineering
- Zero-shot prompting
- Few-shot prompting
- Prompt chaining
- Prompt optimization
- Iterative prompting
- Structured prompting
- Prompt decomposition
Our AI Prompting Techniques guide covers these methods in greater detail.
How AI Terms Relate to Each Other
AI terminology can seem complicated because many concepts are connected.
A simplified relationship looks like this:
Artificial Intelligence → Machine Learning → Deep Learning → Neural Networks → Foundation Models → Generative AI
For language-based systems:
Generative AI → Large Language Models → Prompts → Prompt Engineering
For information retrieval:
AI Model → Retrieval → Embeddings → Vector Database → RAG → Grounded Generation
For autonomous workflows:
AI Model → Tool Use → Workflow → AI Agent → Agentic AI
Understanding these relationships makes the broader AI ecosystem easier to navigate.
Frequently Asked Questions About AI Terms
What is AI in simple terms?
AI is technology that enables computers to perform tasks that typically require human intelligence, such as understanding language, recognizing patterns, generating content, and solving problems.
What is the difference between AI and machine learning?
AI is the broader field of creating intelligent computer systems. Machine learning is one approach used to build AI systems by allowing models to learn patterns from data.
What is generative AI?
Generative AI refers to AI systems that can create new content such as text, images, audio, video, or code.
What is an LLM?
An LLM, or large language model, is an AI model trained on large amounts of data to understand and generate language.
What is a prompt?
A prompt is an instruction, question, or input given to an AI system to guide the output.
What is prompt engineering?
Prompt engineering is the process of designing and refining prompts to improve the quality, relevance, and reliability of AI outputs.
What is a token in AI?
A token is a unit of text processed by an AI language model. Tokens can represent complete words, parts of words, punctuation, or other text elements.
What is an AI hallucination?
An AI hallucination occurs when an AI system generates information that appears convincing but is inaccurate, unsupported, or fabricated.
What is RAG in AI?
RAG stands for Retrieval-Augmented Generation. It allows an AI system to retrieve relevant information from an external source and use that information when generating a response.
What is an AI agent?
An AI agent is a system that can use AI models, tools, and multiple steps to pursue a goal and perform tasks with some degree of autonomy.
What is multimodal AI?
Multimodal AI can work with multiple types of information, such as text, images, audio, and video.
What is fine-tuning?
Fine-tuning is the process of adapting an existing AI model by training it further on a more specialized dataset or task.
What is an embedding?
An embedding is a numerical representation of information that allows AI systems to compare and work with relationships between pieces of data.
What is a context window?
A context window is the amount of information an AI model can consider as context during an interaction.
What is an AI API?
An AI API allows developers to connect applications to AI models programmatically.
What is AGI?
AGI stands for Artificial General Intelligence. It generally refers to a hypothetical AI system capable of performing a broad range of intellectual tasks rather than being limited to specific tasks.
What is the difference between an AI chatbot and an AI agent?
An AI chatbot primarily interacts with users through conversation, while an AI agent can be designed to take actions, use tools, and complete multi-step tasks toward a goal.
Why is AI terminology important?
Understanding AI terminology helps you use AI tools more effectively, understand their capabilities and limitations, communicate about AI clearly, and make better decisions when choosing AI tools.
Explore Promptenza
The AI world is constantly evolving, and learning the terminology is only the beginning.
Continue exploring Promptenza’s AI resources:
- Learn Prompting – Learn how to write better AI prompts.
- Prompting for Beginners – Start with the fundamentals of AI prompting.
- Prompt Anatomy – Understand the components of an effective prompt.
- Prompt Engineering – Learn systematic methods for designing and improving prompts.
- AI Prompting Techniques – Discover practical techniques for getting better AI results.
- ChatGPT Prompting – Learn prompting techniques for ChatGPT.
- Gemini Prompting – Learn how to create better Gemini prompts.
- Claude Prompting – Explore prompting techniques for Claude.
- AI Image Prompting – Learn how to write effective prompts for AI image generators.
- AI Video Prompting – Learn how to create better AI video prompts.
- Prompt Library – Find reusable AI prompts.
- AI Tools – Discover AI tools for different tasks.
- Tutorials – Follow practical AI tutorials.
- Resources – Explore useful AI resources.
Final Thoughts
Artificial intelligence can seem complicated because of the number of new terms used to describe its technologies, models, applications, and techniques.
You don’t need to memorize every term.
Start with the fundamentals: AI, machine learning, generative AI, LLMs, prompts, tokens, context, hallucinations, RAG, embeddings, and AI agents.
As you use more AI tools, the terminology will become easier to understand.
Keep this AI Glossary bookmarked and return whenever you encounter an unfamiliar AI term.
Promptenza helps you understand AI, learn prompting, discover useful AI tools, and become more effective at using artificial intelligence.
Promptenza
Learn to Prompt. Get Better Results.
