Cost per token
The price you pay for each token processed — input and output are usually charged separately.
Think of it like
Like paying by the word instead of by the page.
Example
A model might cost $0.01 per 1K input tokens and $0.03 per 1K output tokens; a 1000-token response costs $0.03.
How it actually works
Different models have different per-token prices. Larger models, newer models, and models with longer context windows cost more per token. Input and output are usually split because output tokens require more compute. Your bill is input_tokens × input_price + output_tokens × output_price.
For product teams
Per-token pricing is the main driver of LLM inference costs — choosing the right model size and prompt compression matters hugely.
For engineers
Price is set by the provider; your job is to count input and output tokens accurately and track spending.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome