what are tokens in AI? the units that decide cost, speed, and how much your model can read

AI doesn't read words — it reads tokens. They're why API pricing, context limits, and speed are all measured the way they are. Here's what a token…

Aliteq
Lena Fischer · AI & Local Compute Editor

The short version

A token ≈ ¾ of a word (~4 characters). Common words are one token; longer/rare words split into several.

The short version

AI reads and writes in tokens, not words or letters — that's the model's native unit.

The short version

Pricing is per token — cloud APIs charge per million input/output tokens.

The short version

Context windows are measured in tokens — how much text the model can consider at once.

The short version

Speed is tokens per second (tok/s) — the standard benchmark for how fast a model generates.

The short version

Rule of thumb: ~1,000 tokens ≈ 750 words ≈ 1.5 pages of text.

Aliteq

Read the full story

what are tokens in AI? the units that decide cost, speed, and how much your model can read

Read the full story on Aliteq