Character-based
0
~4 chars/tok
Word-based
0
words / 0.75
Output tokens are estimated at 1.5x input tokens. The prices in this table are reference values for rough planning and should be checked against official provider pricing before budgeting.
Tokens are the basic building blocks of text used by AI models. They can be words, characters, or even parts of words. One token is roughly 4 characters or 0.75 words in standard English text. AI providers charge based on the number of tokens processed (input) and generated (output).
In standard English text, a token is roughly equivalent to 4 characters. This is a common industry rule of thumb for quick budgeting.
The OpenAI and Anthropic tokenization models generally map 1000 tokens to approximately 750 words. Divide your word count by 0.75 for a precise word-to-token projection.
Standard LLM pricing is measured in units of 1 million tokens. We calculate both input (your prompt) and a estimated output (the AI response) to give a total budget.
Specially picked tools to help you plan better.