OpenAI Token Counter

Count text tokens and estimate cost for listed OpenAI model profiles

Last updated: October 14, 2026

How to use

  • Select an OpenAI model from the dropdown at the top
  • Paste your prompt or text into the input field
  • Token count updates automatically as you type
  • View token count, cost estimation, and context usage
  • Click "Copy Results" to save the information
  • Output tokens are estimated at ~15% of input tokens

Note: Token counting uses the js-tiktoken library for accurate counts. Output token estimates are based on typical usage patterns. For precise cost estimates, consult the OpenAI Pricing page. Pricing is based on current OpenAI rates and may change.

Result actions

Copy or download the tool output, or share a link to OpenAI Token Counter.

Get new tools in your inbox

An occasional email when we ship new free tools. No spam, unsubscribe anytime.

Related tools

Estimate tokens for text sent to an OpenAI model

Paste text and choose one of the model profiles in the dropdown. The counter uses a matching js-tiktoken encoding when it loads successfully. If tokenizer loading or encoding fails, it estimates input tokens as one token per four JavaScript string characters. That fallback is approximate and can differ from the API tokenizer.

The page estimates output length as 15% of the input token count; it does not generate or count an actual model response. Cost and context usage use the values stored in the tool’s model table. Model availability, token handling, context limits, and API prices can change, so verify current OpenAI documentation and pricing before making a budget or API decision.

Text is counted in your browser and is not sent to OpenAI by this tool.