Public price and API response consistency check

Document identifier:

This identifier marks the current version of this document. It is not an API key, an access password or the name of a service feature.

1. Open the model market and look up the model identifier (model ID) and its billing items. Prices follow the live data on that page; the Markdown version of the page is a snapshot taken at publish time.
2. Call GET /v1/models with your own API key to confirm the model identifiers your account can use. For video models, see the catalog and parameters in the video model docs.
3. Send one request, find that call in the console usage details, and compare it with the public price. For text models, multiply the usage by the unit price of each billing item listed in the price table (for example input, output and cache), and mind that the pricing unit is per million tokens; where the table marks tiers (for example cache or long context), use the matching tier. For models billed per task, such as video, the public pages carry no unit price you can multiply directly: compare the task's estimate and usage details in the console instead.
4. Compare with the transparency statement to confirm which categories of billing and log data the platform itself keeps.

This check only verifies that the public information on the site agrees with the applicable API responses and billing records. It does not authenticate the identity of an upstream model, and it does not imply endorsement of iwkey by any model vendor.

If you find that public information does not match actual responses or billing, contact [email protected].

Document version: 1.0
First published: 2026-10-23