Glossary

LLM token cost

A plain-English definition, how it works, an example, and how it is used in enterprise AI.

{ Definition }

What is LLM token cost?

The price of using a large language model as a service, charged per token, a small chunk of text of roughly three quarters of a word, for both the input sent and the output returned.

How it works

Every request to a hosted language model is billed by tokens: the question, any documents or context sent with it, and the answer. Costs look small per request but grow with three things: how many people use AI, how often, and how much context each request carries.

Agents multiply the cost, because one task can mean many model calls, each carrying context. Retrieval-heavy designs that paste long documents into every prompt add more.

Costs can be cut by sending less context (facts from an ontology instead of pages of text), answering exact questions with queries and rules instead of the model, and running models on capacity you plan rather than per token.

An example

Illustrative example

The free AI cost calculator shows an example: 2,000 people, with agents making 5 calls per task, comes to about $66,000 a month at pay-per-token prices. Change the inputs to see your own estimate.

How NexusAI uses it

NexusAI is sold as a subscription with no per-word or per-token bills, and runs on your own servers or a sovereign cloud, so AI cost becomes capacity you plan. Each copilot answer shows the tokens it used.

{ FAQ }

Questions about LLM token cost

What is a token?

A small chunk of text that models read and write, roughly three quarters of an English word on average. Providers bill per token, for input and output.

Why do AI agents cost more than chat?

One agent task can make several model calls, each with its own context, so tokens add up per task, not per question.

Get started

Build the intelligent enterprise

Turn enterprise systems into an intelligent operating layer that understands your business and executes operations autonomously.

Live in production within 90 daysNo systems replacedCloud, on-premise or hybrid