Tokens
Tokens are pieces of text that LLM operates with. They calculate both the context window and the cost of the API. Russian text “eats” more tokens.
A token is the minimum unit by which LLM “sees” text. This is not a letter or a word, but a piece: for English - approximately 0.75 words per token, for Russian it is worse - the Cyrillic alphabet is divided into smaller pieces, so the Russian text takes up noticeably more tokens.
Why does a marketer need to know this? First, billing: API models are charged for input and output tokens. Secondly, the context window is also measured in tokens - a long Russian prompt takes it up faster than it seems.
Practical conclusion: if you drive volumes through the API, count tokens in advance. Over the course of the quarter, I got real amounts through Veo and text models - understanding tokens helps me not to be surprised by the bill.
Related terms
Where is it understood in practice?
Need to set this up on your project?
I analyze metrics, calculate unit economics and collect funnels on real budgets. 30 minutes on call - free.