# Token

Source: https://customlabs.io/glossary/token/
Updated: 2026-09-13

Foundations

# Token

A token is the basic unit of text an LLM reads and writes.

It is roughly three-quarters of a word in English, though the split varies by model and tokenizer.

Usage-based pricing, context-window limits, and latency are all denominated in tokens.

Token counts are the unit every cost or capacity conversation in this space eventually collapses to.

[Insight: What AI actually costs](https://customlabs.io/insights/what-ai-actually-costs/)[AI Cost Calculator](https://customlabs.io/tools/cost-calculator/)

[← Back to the full glossary](https://customlabs.io/glossary/)

## Related terms

[Context Window The context window is the maximum text, measured in tokens, a model can consider at once.](https://customlabs.io/glossary/context-window/)[Inference Cost Inference cost is what it costs to run a trained model on a request.](https://customlabs.io/glossary/inference-cost/)

## More in Foundations

[Retrieval-Augmented Generation (RAG) Retrieval-Augmented Generation (RAG) retrieves relevant passages at query time and feeds them into an LLM's context.](https://customlabs.io/glossary/retrieval-augmented-generation/)[Embeddings Embeddings are numeric vectors that place similar content close together in vector space.](https://customlabs.io/glossary/embeddings/)[Model-Agnostic Architecture A model-agnostic architecture lets an application swap or route between LLM providers without a rewrite.](https://customlabs.io/glossary/model-agnostic-architecture/)[Fine-Tuning (vs RAG) Fine-tuning trains a model's weights on your own examples, changing its behavior directly.](https://customlabs.io/glossary/fine-tuning/)