Count tokens locally, check context usage and estimate request costs with your own rates.
Choose the encoding used by your model and paste the complete text. Enter context capacity, reserved output tokens, input/output USD rates per million tokens, and total planned requests. For daily or monthly forecasts, enter the total requests for that period.
1,000 input tokens at $1/million plus 1,000 output tokens at $3/million cost $0.004 per request, or $4 for 1,000 requests.
Up to 100,000 characters. Algorithm assets load from this site on first use; no model call or text upload occurs. Costs assume identical input and the full reserved output on every request, excluding cache discounts, batch rates, tiers, taxes and other charges.
Why does API usage differ? This counts raw text using the selected encoding. Protocol and model overhead add tokens. Reserved output is a planning assumption, not a prediction of actual generation length.