Skip to main content
The token counter shows an estimated token count and input cost as you type your task description. It helps you predict API costs before running a task.

​
How It Works

As you type in the task creation modal, the counter calculates:
  • Token estimate — Approximate tokens based on character count (~4 chars per token)
  • Input cost — Estimated cost using the selected model’s pricing
  • Character count — Raw character length of your prompt
The counter appears below the description field once the content exceeds 10 tokens.

​
Model Pricing

Cost estimates are based on the model selected in task options:
ModelInput Cost (per 1M tokens)
Haiku$0.25
Sonnet$3.00
Opus$15.00
These are input token estimates only. Actual task cost includes output tokens, cache tokens, and tool usage which vary by task complexity.
Use the token counter to spot overly long prompts. A concise, focused prompt often produces better results than a verbose one.