Changes

After this price change, when requests to the following models (see tables below) in the Singapore and Beijing regions hit the cache, the hit input tokens will be billed as cached_token. The unit price changes from 40% of the input_token price to 20%. Input tokens that do not hit the cache are billed at the standard input_token rate.

Impacts

After the update:

• Beijing region pricing:

• Singapore region pricing:

Learn more about context cache. If you have any questions, submit a ticket to contact us.