PIPELINEGlossary · AI and automation
What is a language model?
A language model is a program that has learned from very large amounts of text which word is likely to come next in a given context. This creates the ability to understand, summarize, translate, and formulate texts — without the model understanding what it writes in a human sense.
01 — Context
Language model (LLM) in detail
The abbreviation LLM stands for Large Language Model. “Large” refers both to the amount of training material and to the number of internal parameters through which the model stores relationships.
A language model has no memory between two conversations and no access to current information unless you give it that access. Both can be added technically — through a knowledge base, through tools, or by including previous messages.
02 — Practice
What this means for small businesses
For businesses, what you provide to the model is more critical than the choice of model itself. The same technology delivers useful answers with a well-maintained knowledge base and arbitrary ones without it.
03 — Questions
Frequently asked questions
Which model is the best?
That changes several times a year and is secondary for most business applications. More important are data protection, availability, costs, and whether you can switch providers without rebuilding everything.
Does the model learn from our inputs?
With business accounts under the appropriate contract, usually not. With free consumer accounts, that's often different — that's precisely where the data protection distinction lies.
04 — Continue
Related terms and services
- Automation
- Knowledge base
- AI phone assistant
- Prompt
- AI agent
- Web accessibility
- CRM system
- AI Overview
- Conversion
- All terms
Related services
Next step
Let's talk briefly about whether this makes sense for you.
A twenty-minute conversation is usually enough to clarify whether a project makes sense. If not, we'll say so — that's cheaper for both sides than an offer nobody needs.