Tokens
In OpenAI documentation, tokens are the pieces of data passed to the model, from text to images. The context window is measured in tokens.
On this page
Source content checked . English translation published 2026-10-27.
In business automation
Tokens are the units the model processes text in, and are not necessarily whole words. Their number affects context size and service usage, so examples should be measured in Hebrew and with the actual input structure.
Examples
These examples illustrate a possible workflow. Adapt the connections and rules to your own systems.
Business process example
A document summarization process measures the size of the instructions, document, and response. Long documents are split or summarized in stages, while retaining a reference to the source so important details are not lost.
Relevant tools
- Anthropic
- Model usage is measured in input and output units according to the model and service. Check the actual usage data.
What to check
- Do not estimate cost based only on the number of words. Instructions, history, tool results, and output may also be included in the calculation.