Alibaba Cloud's large-scale model service platform, Bailian, has announced a price reduction for context caching of certain models, effective from August 26th. Following this adjustment, requests for models where input tokens are cached will incur a charge that is 20% of the standard input_token unit price. For tokens that are not cached, the standard input_token rate will apply.