Token Counter

Count tokens for a text string using the GPT tokenizer. Useful for estimating API costs.

Requires an API key. Run devv login <key> first.

Commands

devv tokens "Hello, how are you today?"
devv tokens "$(cat prompt.txt)"
cat long-prompt.txt | devv tokens

Examples

# Count tokens in a prompt
devv tokens "Explain quantum computing in simple terms"

# Count tokens in a file
cat system-prompt.txt | devv tokens

# Estimate cost before sending
devv tokens "$(cat my-prompt.txt)"

Output

Tokens: 7
Model estimates:
  GPT-4:     ~$0.0002 (input)
  GPT-3.5:   ~$0.00001 (input)

Pipeline Usage

Pipeline ID	Description
`token-counter`	Count GPT tokens

devv pipe exec "token-counter" "Hello, how are you?"

Behavior

Takes text as input, returns token count and cost estimates.