Your AI SaaS business, ready by this week!
Try demoYour AI SaaS business, ready by this week!
Try demoTokens are the small pieces of text an AI model reads and writes, roughly four characters or three quarters of an English word each, and the unit most AI providers bill by.
Before a model sees text, a tokenizer splits it into tokens: whole short words, parts of longer words, punctuation and spaces. A rough rule for English is 100 tokens per 75 words; other languages often need more.
Providers charge separately for input tokens (your prompt, the conversation so far and any attached documents) and output tokens (the answer). Output tokens usually cost several times more.
Because the whole conversation is sent again with every message, long chats get more expensive per message. That is why AI products cap history, summarise it or charge more for long contexts.
Aikeedo counts the tokens of every request and converts them to credits at the rates you set per model, so users see a credit balance instead of token maths.
Discover more terms connected to this one
AI credits are a unit an AI SaaS sells to its own users and converts to tokens, images or seconds of audio at rates it sets, as opposed to the tokens AI providers bill by.
The context window is the maximum number of tokens a model can take into account at once, covering the instructions, the conversation, any attached documents and its own answer.
An LLM (large language model) is an AI model trained on very large amounts of text to predict the next token, which lets it write, summarise, translate, reason and hold a conversation.