Pricing

Read the current token rates

Rates are quoted in USD per one million tokens. Input, cached input and generated output have separate prices. The table below shows current rates for models marked Available.

• View published pricing →See current availability, announced offers and standard rates.

Calculate a request’s cost

Cached input is part of the total prompt token count. Subtract cached tokens from prompt tokens before applying the uncached input rate, so the same input is not counted twice.

uncached input = prompt tokens - cached input tokens

cost = (uncached input × input rate
      + cached input × cached input rate
      + output tokens × output rate) / 1,000,000

Arnict totals the categories and rounds once per request to its accounting precision. The request record in Usage is the reference for the recorded amount. Requests with no confirmed usage are not charged.

Billing and usage records

The dashboard shows confirmed requests, token counts, cost and your available balance. Paid model requests need a positive balance when they start, then deduct their confirmed cost after completion.

Add purchased credit in Billing. Checkout is hosted by Stripe on its own site, and purchased credit does not expire. Check Billing and the purchase terms for the current first-top-up bonus offer and eligibility.

Check the request record for the rate and confirmed token usage used to calculate its cost.

• Review Usage →Inspect recorded requests and costs.
• Open Billing →Check the payment options currently available to your account.