Tool descriptions are full of shorthand. This glossary covers the terms you're most likely to run into while browsing the directory, with a sentence or two on each. It isn't a textbook. The goal is to help you read a listing and know whether it's relevant to you.
The basics
AI model
A program trained on large amounts of data to produce an output, such as text, an image, speech or a prediction. Most tools in the directory are products built on top of one or more models.
LLM (large language model)
A model that works with text. It predicts what words should come next, which turns out to be enough to answer questions, summarize, translate and write code. Chat assistants and most writing tools run on an LLM. You'll find providers and hosting services in the LLM Platforms category.
Generative AI
AI that creates new content, such as text, images, music or video, instead of only sorting or predicting. Image generators and music generators are examples.
Prompt
The instruction you give a model. A good prompt says what you want, gives context, and states the format you'd like back. Tools in Prompt Engineering help you write, test and store prompts.
Hallucination
When a model states something that sounds right but is false, such as an invented quote, link or statistic. It's the main reason to check important output yourself.
How tools are built
Fine-tuning
Training an existing model a bit more on your own examples so it behaves in a specific way, for instance writing in your brand's tone.
RAG (retrieval augmented generation)
A method where the tool first searches your documents for relevant passages, then gives them to the model so its answer is based on your material rather than only on what it learned in training. If a tool says "chat with your PDFs", it usually uses something like this. See RAG Tools.
Embeddings and vector databases
An embedding turns text, or an image, into a list of numbers that captures its meaning, so similar things end up close together. A vector database stores these and finds the nearest matches quickly. They power semantic search and RAG. See Vector Databases.
API
A way for one piece of software to talk to another. When a tool offers an API, developers can plug it into their own apps. Pricing for APIs is usually based on usage.
Open source
The code, and sometimes the model, is public so anyone can inspect it, run it themselves and modify it, subject to the license. It isn't automatically free to run, since you may need your own hardware.
Agents and connections
AI agent
A system that doesn't only answer but takes steps toward a goal: searching, calling other tools, writing files, and deciding what to do next. Agents range from careful assistants to risky autopilots, so look at what they're allowed to do. Browse AI Agents.
MCP (Model Context Protocol)
An open standard that lets AI assistants connect to outside tools and data, such as your files, a database or a calendar, in a consistent way. An MCP server exposes a tool to the assistant. See MCP Servers.
Chatbot
Software you talk to in a conversation. Some are general assistants, others answer customer questions for a single company. See Chatbots.
Automation
Setting up tasks to run by themselves when something happens, for example "when an email arrives, summarize it and add a row to a spreadsheet". See Automation.
Media and language terms
Text to speech (TTS)
Turns written text into spoken audio. See Text to Speech.
Speech to text, or transcription
The reverse: converts recordings into written text. See Transcription.
Voice cloning
Producing speech that sounds like a particular person's voice from a sample. Only use it with the speaker's clear permission.
OCR (optical character recognition)
Reads text out of images, scans and handwriting so you can search and copy it. See OCR.
Upscaling
Increasing the resolution of an image or video, with the model filling in plausible detail. See Image Upscaling.
Dubbing
Replacing speech in a video with another language while keeping timing, and sometimes the original voice. See AI Dubbing.
Text summarization
Shortening long text, audio or video to the key points. See Text Summarization.
Pricing words
Free, Freemium, Paid, Free Trial and Open Source are explained in Pricing labels explained.
Not finding a term?
If a listing uses a word you can't work out and the tool's website doesn't explain it, tell us. We'll add it here.
